Speed Work


Experience From - Sean McDonald, Matt Mahoney, Mike Bate, George Beinhorn#1 , Blake Wood#1 , Fred Vance , Douglas Spink, Matt Mahoney , George Beinhorn#2 , George Beinhorn#3 , George Parrott , Blake Wood#2 , Damon Lease , Mel , Blake Wood#3, Steve Pero Bill Ramsey , Rich Lacey , Karl King , Scott Weber , Al Zeller , Scott Weber#2 , George Parrott , Daniel Dreyer , Carey , Pete Petri , Brian Pickett, Shawn McDonald#2, Jeff Riddle, Shawn McDonald#3,

Sean McDonald

The question is: how can I increase my cruising speed without having to run 10k's as recommended by some runners?

There are many ways to increase your cruising speed. They all boil down to stress followed by rest leading to a variety of changes, giving greater running speed with the same effort. Does this mean going out and just hammering a training run once each week ? I don't think so. The key is to gradually build up the volume and pace of the harder efforts during the "speed" session, and perhaps gradually decrease the rest interval. There are many types of speed play to choose from:

Running on a route (roads, trails, etc.) where you occasionally pick up the effort/pace for short time periods (often 30 seconds. to a few minutes), with recovery between the pickups. You can decide when to run the harder pickups based on terrain and/or how you feel (i.e. is your breathing and heart rate recovered down below a normal "long" run effort). This type of workout can be done anywhere, alone or with a group.

Hill Repeats:
Running a hill of length 150 to 800 or so meters, with a grade of a few percent (can be on trails, fire roads, paved roads), running up the hill at about 10k effort, and recovering at an easy jog down the hill.
running set distance and/or time on flat terrain (track or pavement) at a fast pace, with stipulated rest periods of slow running between each fast interval, can have all the intervals the same length/time, or different length and/or paces - many people find this type of work out is easiest done with other people. Will help to develop a smoother running form, and ability to handle anaerobic work.
Hill Fartlek:
Running a given route (on rolling hilly terrain, dirt roads or pavement are best) by running several of the up hills at a harder effort, with recovery going down the following down hills. Like the hill repeat workout, this will develop running strength, and a more efficient uphill stride.
Tempo Run:
A run where you do a couple or three warmup miles at normal pace, then do a period of faster running (typically at 10k-marathon pace) followed by a cool down at normal pace. This teaches one how different effort levels "feel" and works on increasing the running speed of your anaerobic threshold.
These are some of the speed runs that you can put into your schedule to improve your cruising speed. It will take a few weeks before you see any changes though so be patient. If you are just starting (or getting back to) speed training, I would do one session every other week. Start with just a few of the intervals/repeats, say as a Fartlek, do: a 2 mile warmup, Fartlek pickups at your 10k pace, 6 minute total of pickups (30 seconds. to 2 mins. each) recovery time after each pickup then a 3-4 mile cool down. Choose a route with fairly flat terrain for this run

Another good beginner workout is hill repeats, substitute 5 x 1 mins. uphill at 10k effort for the above Fartlek pickups. The Fartlek and short hill repeats are easier on the body than longer/faster intervals on a track, so do these early on in your speed program, and after several sessions graduate to intervals and tempo runs.

Have fun with these speed runs, do a good warmup before, cool down after, and stretch after the session. Vary the # of hard efforts, terrain, pace, it will be more fun that way. It is very feasible to do some of the workouts on fire roads or trails (as long at they are not super rocky), just watch out that you don't push the up hills too hard (very easy to do). As an added benefit, this will improve your trail agility, since you are going faster than normal and will have to react to obstacles faster. The day after the speed run (at least, or maybe the two days after) do just a short, easy run, or a short cross training session. Give your body a chance to recover, and build up based on the stress you put on it in the speed session.

It may be easier to do the speed sessions on alternate terrain (non-pavement) if you wear a heart rate monitor, but then you need to know the range you are shooting for during the harder periods, and what you want to recover down to during the rest times.

Good luck to all with your running. Doing a faster training session on a regular basis can help you to run races faster, and stay ahead of the cutoffs. Just do it sensibly, and rest after.

Note: As a reference to speed work, check out the Competitive Runners handbook by Bob Glover, or Jeff Galloway's book, or most issues of Runners World. You can probably find these at your local library.

Matt Mahoney

Robert Flanigan wrote:

"I am new to this scene and am looking for a few examples of how you do speed work, If you do it at all. Also, do you taper before a big race?"

Yes, it is about a third of my total mileage. (The other 2/3 are short races and long slow runs). Usually it is a Fartlek run where I run hard for a couple of minutes, jog or walk to recover, and repeat several times. I don''t time or measure my intervals but they're typically 15-30 seconds at 5:00/mi or 1-3 minutes at 6:00/mi or something like that. I go by effort, saving the hardest sprints for last.

Mike Bate

As regards speed work I once read an article by Bruce Fordyce where he seemed to advocate a policy of run fast, run long then taper as phases in training: i.e. a period where speed work is dominant (say 3 sessions a week) for a month, and then build up over a month to maximum mileage holding at as a plateau for 2 months to go to 1 month to go from the race, with less concentration on speed, but doing a handful of long runs. Then taper using one or short races for sharpening.

Has anybody any experience of this 'phased approach'? does it work for the common mortal?

From reading other postings the ones who seem to be most consistent, and with fewest DNFs are the ones who adopt a fairly stable training regime year round. It seems to me that all of the changes in training mode are the risk points for injury, so I would hazard a guess that the consistent trainers are least affected by injury.

But maybe it does need the building up to a crescendo to achieve best performance (and therefore risking a DNS)

I don't live near any hills so I need speed work if I'm going to do hilly courses like Leadville or Barkley. Speed work gives you the endurance for climbing and the leg speed for descents.

I don't normally taper before a big race because my mileage is already pretty low (15/week) and fairly constant throughout the year whether I'm training for a big race or not (and I usually am). I normally don't run for about 4 days before or after a 100 mile race, but I usually go that long between runs anyway.

For 100 miles, your longest run should be 1-3 months before your race. In the 100's that I've finished (4 of 7), my long run was 70 to 100 miles (another ultra). In the 3 I didn't finish, it was 26 to 57 miles.

Most ultra runners train about 50 mi/wk with little or no speed work and no runs over 50 miles in preparation for a 100. Also, most of their runs are for aerobic conditioning, with one long run (20-30 mi) every week or two (which is more than I need; every month or 2 will do). I like to bike, swim, and lift weights, but if you would rather just run every day and you don't get injured, then maybe high mileage is for you.

George Beinhorn #1

How to run faster. Pat Wellington raised the question a while back, and it set me to thinking. What I've figured out is that in order to run faster, you have to "do the math." (You've got to do the work, but that comes later.)

The math, at least, is pretty simple. These are some of the factors that determine how fast you can finish an ultra. But first let me propose a simple axiom:

Ultras are not won on the uphills. Everyone goes slow uphill, and the differential between the uphill speed of the upper and lower third of the pack is not great. But ultras can be "lost" on the uphills, if a person is poorly prepared for climbing.

Ultras are won on the downhills and flats, where the upper third of the pack goes a "great deal" faster than the "caboose." Speed is the killer.

Okay, how can you improve your finishing times? Here are some suggestions, based on doing the math:

  1. Change your run/walk ratio. Simply put, the more you run, the faster you finish. How can you train to be able to run more and walk less? I've been able to think of three ways:

  2. Running speed. If these methods don't appeal, you can simply change how "fast" you run. Don't change the proportion of walking, just floor it during the running segments. This obviously requires that you do some speedwork.

Speedwork has an unfortunate reputation. Yes, it is painful, but if you've got a certain twisted type of mentality, it's also fun. After the first 2-3 speedwork sessions, the "bad pain" begins to turn into "good pain." Also, speedwork carried the fringe benefit that it energizes you all day long.

What kind of speedwork should you do? Eric Robinson does fast quarters. At least one world-class masters marathoner does fast 800's exclusively. As a 55-year-old poop-along, I enjoy fast quarters, with a few halves and miles thrown in for endurance. I jog two minutes very slowly between repeats, and I'm toying with doing an occasional tempo run of 25-30 minutes. On expert advice, I never do more than 3 miles of fast running per week, except for tempo runs. I try to do one speedwork session per week.

When I began doing speedwork at age 53, I ran my first mile repeat in 7:42, painfully. After 16 weeks of speedwork, I ran 10 miles at 7:00 pace. Quite a difference--and speedwork helped me in my ultras, too. I felt less sore, and the running segments felt just as easy, even though I was running them faster.

I've neglected speedwork for the least two years. I'm just getting back to it now, and I thought I'd share my thoughts.

Blake Wood #1

George Beinhorn writes:

"Ultras are not won on the uphills.... Ultras are won on the downhills and flats, where the upper third of the pack goes a great deal faster than the "caboose.""

How true this is! This past summer, I made a real effort to run lots of hills in training, and to run them fast - particularly downhill, in order to build up my quads. I could really feel the difference at Hardrock - for the first time in my three finishes there, my quads were in reasonably good shape all the way to the end, and I was able to run the downhills fast the whole way. This made a big difference - the difference between going fast and slow uphill (a steep hill, where everyone walks) is probably substantially less than a factor of two, but the difference between walking and running downhill is probably a factor of 3-4. Certainly, there is even a difference of more than two between "shuffling" and running downhill.

Fred Vance

George wrote:

"Ultras are not won on the uphills."

In 1991 when Andy Jones set his North American 100K race with the controversial use of a heart rate monitor, it was his stated game plan to stay at a heart rate that kept him just within his aerobic limit. Andy down-rated slow twitch muscles in an interview at the time.

That's one extreme of the sport, but it didn't include any uphills. At the other extreme, Mark Williams became the first and only person to finish Barkley.

Where Andy was laying down a long string of sub-six minute miles, Mark's average pace was probably something like 30 minutes per mile. Both are winners. One is a master of speed and endurance, the other a master of endurance and perseverance.

Uphill may be insignificant or it may be all important. It depends on the course. It seems to me though, that people that can run or even walk faster uphill have a huge advantage in many trail ultra events.

I work on speed during my noon runs out of the necessity of trying to get in as many miles as possible in 50 minutes, but give me five or ten hours in the mountains on a weekend any day over speed work.

Douglas Spink

How many people integrate a structured speedwork component into their regular training for ultras? I've seen research suggesting that track-based intervals of 400-1200 yards done in a timed environment can really help to build leg speed on longer runs.

I am a slow ultra runner, and have never really been fast at distance. As I put together my 1998 training plans, I am wondering if I should heed this research and spend a day a week at the track belting out quick 400s. Or, am I kidding myself and I should just accept the fact that I am a big, slow, persistent-as-hell ultra participant?

I am thinking of more than fartleks here-my idea is actual, track-based speed workouts. I am not suggesting speed workouts as a replacement for distance, but rather as a compliment to longer work.

Matt Mahoney

Bill LaDieu wrote:

"Any thoughts how early to start speed work before a planned ultra?

I have a thought. Do speedwork once a week, year round. Make it fun, so you'll want to do it. You don't need a track and a stopwatch. Instead, try "Race you to that tree!" A lot of my intervals are on dirt. Just run hard 1-2 minutes, walk, repeat, and save the hardest for last. Try them at the end of a 10-15 mile run.

I discovered the value of speedwork several years ago when the local running club put on weekly track meets. The events were 100 yards, mile, 440, 880, 220, and 2 miles. I would enter tham all. 5 weeks later I unwittingly took a minute off my 5K time. Hey, that was easy.

For ultras, the value of speedwork is in downhill running, especially because there are no hills where I live. There is absolutely no way I could finish trail 100's on my mileage (15/week) if half of it wasn't speedwork. (I run twice a week, the other run is either a race or a long run. On other days, I bike or lift weights).

4 miles of repeats. Not particularly fast, but I was running barefoot in soft sand on the beach at high tide, occasionally knee deep in surf. There was a storm off the coast of Florida that made the water rough to the delight of the surfers, but the weather here was perfect, 70 deg. and sunny.

George Beinhorn #2

I'm very new to speedwork. My speedwork coach was a champion N. Calif. 60-64 group roadrunner who was "very" unsystematic in his approach. He just went to the track (with a group of his buddies, including me) and did whatever he thought up on the spur of the moment. "Let's do a pyramid--four quarters, one half, one mile, and back down."

Then we just took off and did whatever we could. No prescribed pacing. Nothing. Carl did speedwork year-round, and it worked for him. Of course, he was retired and spent winters in Hawaii, too bad.

It worked for me. Carl's only rule was never do more than 3 miles of fast running in a speedwork session. He thought Jeff Galloway's recommendation that marathoners work up to 13 repeat miles was just crazy, because most marathoners will try to do it too fast and fry themselves in the process. I know two runners who experienced exactly that.

Sorry to offer so little help, but you need to hear from someone much more experienced than I, if you want to be systematic about it.

George Beinhorn #3

"Uphill may be insignificant or it may be all important. It depends on the course. It seems to me though, that people that can run or even walk faster uphill have a huge advantage in many trail ultra events."

I agree with you completely. Uphills can trash you so badly that you're dead on the flats and downhills. Before WS97 I did a lot of weight training. During the race I felt just hugely strong on the uphills of the canyons, and really bombed up them in a big exhibit of joyful power. I paid for it dearly, as I think it's exactly what brought on the rhabdomyalgia that forced me to DNF.

"I work on speed during my noon runs out of the necessity of trying to get in as many miles as possible in 50 minutes, but give me five or ten hours in the mountains on a weekend any day over speed work."
Either/or wasn't what I had in mind. You've added valuable input to the dialogue. Thanks.

George Parrott

Yes, year around speedwork is very good, but perhaps with more of a pattern and progression in it than what you (George B.) are noting from Carl Ellsworth.

I love Carl, but he is simply "intense" in what he tries to do and not as "analytical" as he might be...IMHO, of course.

Chris and I just got back from Hawaii yesterday, after staying with Carl and Tsuru there for the Honolulu Marathon--we also did CIM the week before. For road and even trail ultras, marathons can be a "kind of speedwork" too, and so can regular (every 3 weeks or so) 10k road events. I don't agree with Carl at ALL on his "max speed volume" formula, though I think 3 miles is "usually good." For marathon preparation I try to frequently move the CHIPS to 4-6 miles of speed effort in various patterns, e.g.

6 x 1 mile or...
4 x 1.5 miles or
3 x 2 miles or...
other variations....with typically 200m or at most 400m jog recoveries.

These short repeat sessions should be run at about "10K RACE effort" with the last one .....all out.

Blake Wood #2

Douglas Spink asks:

"How many listers integrate a structured speedwork component into their regular training for ultras?"
I started doing this last summer, and I feel like it helped. Not so much because it improved my speed, but because it was my most intense workout of the week - I was really tired afterwards. I typically did this once a week - I'd run to the local track (a 9 mile round trip) then do four one-mile timed runs, jogging a quarter mile in between. Sometimes I'd substitute four quarter miles for the final mile, but generally kept constant the total of four miles of timed running, for at total of about 14 miles for the workout. I always did this in the morning - in the spring and fall I usually watched the sunrise from the track, and was home by 7am. On a good day, I could keep all four one-miles under 5:40. Due to the altitude (7200'), I could never seem to go much faster than this - my quarter miles were rarely under 75 s, limited by how much oxygen I could suck in.

Damon Lease

A few years ago, there was an article in RW about a workout the author called "Yasso 800's." The name came from Bart Yasso, a RW staff member who has run about 2:40 for the marathon, and has done a number of ultras, including Badwater.

His workout was for the marathon, and I've used it with great success there, and I think it would apply well to ultras. It is a really tough workout, one that left me shot for days afterward.

Essentially, you do 800M repeats with 50% active rest, building up to ten of them. The key is that this workout also is a great predictor of your marathon time.

You should be able to take your fastest average time for the ten 800s, in minutes and seconds, and then that will be your predicted marathon time in hours and minutes.

I used this workout for six months when shooting for a sub-3:00 marathon a few years back. I ran 2:57:35, ran a sub-18:00 5K along the way, and four weeks after the marathon, I just missed my 10K PR as well.

I started about 7 months before the race, doing 6 800s. I averaged 3:18.6 that day. I built up to ten 800s by the fourth time I did the workout, and I averaged 3:11.1 that day. While maintaining my mileage at a high level, averaging 70 mpw, I dropped to 2:57.4 average 5 weeks before the marathon.

I did the workout one final time 2-1/2 weeks before the marathon, after I had started my taper, and I averaged 2:51.1.

I have never found a speed workout that combined the need for speed and strength the way that this one did. As I followed this program, I could feel my speed increase, my strength increase, and my confidence increase.

The day that I actually crossed into a sub-3:00 average for ten is etched clearly in my mind. I was doing this workout alone, it was cool and the sun was setting. It seemed as if I was flying, all alone in the darkness, and I felt like I could do anything that night.

I can't say exactly how the workout would translate to ultras - my fat body isn't prepared for the intensity of that workout right now. But, my guess is that it would be of tremendous benefit if the intensity and need for recovery didn't interfere with your ability to do requisite long runs.


I do that Yasso 800 workout about 2-3 weeks before a marathon to give myself confidence and an accurate, reasonable picture of what kind of pace I can hold for 26.2 miles. For me, it has been very accurate, within about 5 minutes, so long as I don't get too excited in the beginning and start out too fast.

I can't do the same track session over and over, though, or I get bored. Whatever you do, the intervals of work (hard) should be over 2 minutes or you won't reach your VO2 max. 400s are fun and they help with pacing and developing proper form, but it is best for endurance athletes to do longer intervals. In Run Smart, Run Fast, it is also recommended to do all out 100 meter sprints now and then to build up leg strength, but that author (and I haven't remembered an author's name yet, so why should I start now?) was focusing on 5K-10K and a little bit of marathon.

Blake Wood #3

Dennis Herr questioned my numbers, and made some good points in response to my posting, which I thought were worth passing along. I had posted that downhills were more important than the uphills, because the difference between going fast and slow on an uphill was less than a factor of two, but on the downhills it could be a factor of 3-4. An example of the numbers I had in mind are:

Imagine a 10 mile leg over a pass with about 3000' climb (that is, very steep - I had Hardrock in mind). On the five miles up, feeling good, one could probably do 2 mph=2:30. Feeling bad, one could probably still do 1.5 mph=3:20. On the way down, feeling good, one could run, say, 10' miles = 0:50. With trashed quads one would be force to walk, say, 25' miles=2:05 (only a factor of 2.5 slower). That's a 0:50 differential on the uphill leg, but a 1:15 differential on the downhill leg. (Sorry for the mixed units - I tend to think of running in min/mile and walking in mph.)

Dennis commented that although there may be a greater difference between feeling good and bad on the downhill, runners spend a much longer time going uphill, and so small differences in pace can make a proportionally larger difference in elapsed time. In view of my numbers above, I agree that to have most of the time be lost on the downhill, it is necessary that the difference between feeling good and bad on the downhill be great - basically the difference between running and walking. If it's merely a difference between walking fast and slow on the uphills and running fast and slow on the downhill, most of the time could easily be lost on the uphill.

Steve Pero

I find that somewhat high mileage (70mpw) combined with fast hill repeats twice a week will work for me. I say "will" because I am a neophyte ultra runner, having run only one 50, but I know from my past experience that running these workouts will give me my breakthrough someday.

What I do is on one day I run approx 12 reps of a hill that takes about a minute to crest hard, say maybe 10K effort, then 2 days later I do a workout on a hill that takes about 3 1/2 minutes to crest. I run the 1st 3rd hard (sort of like a sprint), walk the next 3rd, then attempt to sprint the final 3rd. I then run fast down, rest a little bit and repeat this 6 times. This way I get my hills and speed all at the same time, plus I like running hills and hate running speedwork. Occasionally, for a break, I'll run a hard 10 miler instead of the 2nd hill workout.

These 2 workouts combined with long runs on both Saturday and Sunday "will" work for me...and could possibly work for You.

Bill Ramsey

Scott Eppelman wrote:

"I am looking for suggestions & guidance - how often? how fast? what kind of workouts? what about recovery?"
I didn't incorporate speed training into my race preparation until meeting and befriending Earl Towner over a year ago. He has top ten finishes at both Leadville Trail and Western States and coaches high school cross-country. When asked if he would be interested in training me for Angeles Crest last year, he agreed. I had two previous finishes at AC100, 29:02 and 25:26, and wanted to go under 24 hours. About two months before the race, he had me doing track workouts once a week on Wednesday night. At the same time I continued building my mileage from 70 up to 95 miles a week. I began my track work with longer distances and fewer repetitions. My track workouts were as follows:

Week 1: 2 X 2 miles at 6:50 mile pace
Week 2: 4 X 1 mile at 6:30 mile pace
Week 3: 8 X 880 yards at 5:50-6:00 mile pace
Week 4: 16 X 440 yards at 5:30-5:40 mile pace
Week 5: 22 X 220 yards at 5:10-5:20 mile pace
Week 6: 6X 220 + 6 X 440 + 6 X 220 at above mile pace
Week 7: Power strides
Week 8: RACE

I questioned Earl about the utility of 220's for someone preparing for 100 miles. He responded with, "When it's late in the race and every other muscle is struggling to work, your body will use those muscles." On race day last year, I ran into some problems with my feet between 30 and 50 miles and fell almost an hour behind sub-24 pace. When I finally got my feet straight at Chilao (52 miles), I picked up the pace considerably. I cruised the second half of the course, a tough feat to accomplish on the AC100 terrain, and finished in 22:57. Earl was waiting at the finish line to embrace me at 3:57 AM in the morning...he obviously knew something I didn't until after that moment. I distinctly remember running the last two miles feeling like I was easily under 7 minute pace...it was incredibly exhiliarting. Speed work pays dividends!

Rich Lacey

Janet Whitesell, George Beinhorn, and Kevin Ash, commenting on my "have your cake and eat it too" posting (ultras and speedy short racing are compatible), discussed three perspectives on the virtues of speedwork in training for ultra runs.

Janet testifies that she set PRs at 5K, 10K and 50 miles in the same year and adds that "speedwork" should embrace intervals, tempo runs, and short races. [That's central because of the importance of increasing the lactic acid threshold...as running Research News articles stress...].

George observes that "the longer you can actually run, the more easily you'll bring down your finishing times," adding that this depends mainly on total training mileage, but also number and length of long runs. So for faster times in ultras, George does not rule out speed training but puts his money on endurance training. (Janet thinks that both speed and endurance training sessions can be fit into a 7 day week, but George would presumably substitute more long endurance runs for the speed sessions in his own strategy.

Evidently George, unlike Janet, wouldn't expect to set a PR in a 5K or 10K, but would stick to ultras. Both Janet and George might set PRs at 50 miles using their separate approaches.) George has nothing against speedwork, then, but apparently believes its effect on ultras, especially long trail ultras, is minimal compared to endurance training.

Kevin Ash responded to George's point by saying that most people can't actually run an entire ultra [depends on how long: 50K is easily runnable for most people...RL] But Kevin believes that "speedwork is key to increasing the time spent actually running during a race."

Then he adds what may be the crucial points: "If I end up walking, it is because I hurt like hell due to lactic acid build-up. The ability to postpone the point where lactic acid begins to accumulate is an important aspect in this situation. Speedwork (tempo runs, and intervals at your lactate threshold in this case) will increase the point where you begin to accumulate lactic acid. I also believe it postpones the point where you begin to develop lactic acid during an ultra. VO2 max is another benefit of speedwork which will increase the time and speed you are able to actually run."

I think the research reported in Running Research News and Peak Performance supports Kevin. Also, an Ultrarunning article a few years back on the 24 hour race between Setnes and Possert illuminates the physiological factors involved in the proportion of walking and running -- particularly the effect of branched chain amino acids (BCAAs) and the tryptophan-seratonin connection which appears to underly extreme fatigue.

I have read that Ann Trason is known to run enormous numbers of repeat intervals on the track at a fast pace. If true, that would also reinforce what the research and personal testimonies (such as Janet's) suggest namely, that she and Kevin are correct that we can indeed have our cake and eat it too -- if we train scientifically.

Karl King

My experience with becoming slower suggests that the easiest way to become slow is to drop the intense work from one's training.

Some history: I started running at age 42, so missed my fastest years. My 10K PR is 44 minutes, so many would say I never was fast. With that level of "speed", cutoffs at ultras were not a concern.

After developing some tendonitis from road racing I had to delete the racing and quality work from my training. At a relaxed pace I could run any distance up to 100 miles with no serious problem.

Eventually the lack of intensity in my training slowed me down to the point where I have to be concerned about cutoffs at the more challenging races ( Leadville for example ).

By treating the root cause of the tendonitis, I've recovered some ability to put some intense work back in the program. The good news is that it doesn't take much serious work to get back some speed. In the space of five brisk runs of five miles ( over the space of 2 weeks ) I took 7 minutes off my 5 mile time. What was surprising was that the effort on the runs didn't feel that much different one from the other; I pushed on all of them. But the speed was markedly different as my body built up the enzymes to extract energy.

If you want to get some speed back, work on your flexibility, especially in the hip area, and make it a point to put in a couple sessions per week where you push lactate threshold for 20 minutes. You can do that on the trails, on the treadmill, on the track, whatever you like. You can do it with intervals or continuous runs; it really doesn't make a big difference so long as you accumulate 20 minutes of hard running in each session. You can also embed such work in your long runs. If you are lucky enough to live near a hill that takes 20 minutes to run up, there you go - just work hard for 20 on that hill.

Many beneficial components in the muscles are prompted to grow when the acidity in the muscle fibers increases. That implies running at or faster than lactate threshold. If you only run slow, very aerobic miles, you won't establish the acidity in the muscles which prompt the growth for speed. If you don't use it, you'll lose it.

Scott Weber

What a pleasure to get Ray the K and Nick Marshall's posts on the subject of Marathon vs. 100 K times. That's a whole lot of running and studying the sport these two have. They know what they're talking about as well as anyone out there.

I won't try to second-quess Charl or put words in his mouth, so I'll ask for clarification on the subject from Charl next time I communicate with him. Should be soon. He's back in South Africa taking care of business.

Strength vs. Speed. Kouros, Noakes, and most students of the sport know that "you can't hide from speed". The faster the 10 K, the more natural talent, the more potential. Charl states that the 10 K is a good barometer. I can't agree more.

Speed + Strength, enhanced by proper training, nutrition, environment, etc. is what will rewrite the record books. Cleverly picking one's parents is what Fordyce has called the determining factor as I recall from one of the Comrades tapes. Who was it who said "The Best Ultramarathoners in the World have not yet run a Ultramarathon..."

Al Zeller

Ray K wrote:

"Regarding Marathon times I rember thinking Charl was off a bit on this one as well as Nick did. Years ago, when I thought such things were important I did similar research. At one point in 1981 I thought I was the slowest marathoner to even come close to 7 hours for 100K. As a 2:48:59 marathoner I ran 5:31 50 miles 7:05 100K. I guess that would be 2 x 2:49= 5:38, oops, I guess I did not have room to add 20 minutes, but actually set a new PR at the marathon enroute, about 2:42, but 2:42 x 2 =5:24, so I had 7 minutes to add. In 1983, with a 2:41 26.2 PR I ran 5:29 50 miles/ 7:04 100K. Later, maybe 1985 I ran a new PR 26.2 2:37:19 (besting a 2:40:12) enroute to a 50 mile/100K of 5:35/7:12. I always thought I needed to go under 2:35-enroute to go under 7 hours at 100K, and that with no second half dificulties. I always thought I had good speed, and had a 34:37 (530ish per mile) 10K, That was pretty fast but certainly did not equip me to go under 7 hours. It was a long time ago, but I believe I did break 6 hours for 50, before the 5:31, while still having the 2:49 marathon Pr. The biggest problem I see with ultrarunners is a tendency to be too conservative. 7 of my best marathon times are still enroute to 50 miles or 100K, I was never convinced of the need to start slowly, and if I ever get back in shape I will blast away gain. Sorry if all these numbers, and actual ultra stuff has offended anyone (you should see my 100 mile data) but sometimes we among the old relic brigade have to say something based on actual experience, after all what else are ultras for?"
I'm with you on this one, Ray, and, of course, I saw you run most of these races. As a 2:50 marathoner I ran a 5:55 50 mile (with a 3:01 split). Then as a 2:47:45 for 26.2 ran a 5:47 50 mile. I never got quite as close to 7 hrs as you did, tho.

That said: Ray the K was, and still probably is, the most competive runner I've ever seen. Going out with him resulted in one of two things: a PR or a DNF. Nothing in between. This doesn't hold for Mr. K, who always finishes.

Scott Weber #2

Paging through Tim Noakes' book, The Lore of Running, I read the following (some times/data are outdated, but it was valid at the time of writing the book):

Eleanor Adams: (p. 356) Adam's best times for the different distances are as follows:

10 km - 36:04; 16 km - 57:36; 21.1 km - 1:18:21; 42.2 km - 2:48:23; 100 km - 8:04:48

Like Trason, there's nobody tougher or smarter about training than Eleanor Adams.

Page 358-Lore of Running: Training Methods of the Elite

"Virtually all great runners achieved success at shorter distance races before gravitating to marathon and ultramarathon races. Kolehmainen, Nurmi, Zatopek, Peters, Edelen, Clayton, Hill, Shorter, de Castella, Salazar, Jones, Lopes, Temane, and Waitz were all excellent track or cross-country exponents before they achieved success at the longer distances on the road, especially in the marathon.

Similarly, the great ultramarathon runners, Hayward, Meckler, Ritchie, Fordyce, Rowell, Kouros, and van de Merwe have all run fast over distances from 10 to 42.2 km.

This evidence proves beyond a doubt that the faster the athlete at short distances, the greater that athlete's potential in the marathon and ultimately in the ultramarathon, as also shown in scientific studies.

This truth was again confirmed in the 1984 Olympic Marathon, won by 1984 World Cross-Country Champion Carlos Lopes, who 2 months before the Games ran the second fastest 10 K ever (27:17:41) Second place in the Olympic Games went to John Tracy, also a former World Cross-Country Champion, who was running his first-ever marathon. Alberto Salazar ran his best marathon when he was training for 10 K on the track, and Steve Jones set his 1984 World Marathon Record in his first marathon, which he ran after training specifically for track and cross-country racing.

Thus it comes as no suprise that Matthews Temane, who has run the fastest mile at altitude, holds the world 21.1 K best; that Bruce Fordyce has the fastest mile, 5 K, and 10 K of anyone running the Comrades; that Fritth van de Merwe and Eleanor Adams are the fastest female marathoners currently competing in the short and long ultramarathons; and that Yiannis Kouros is the fastest marathon runner competing in the long ultramarathons.

The truth is that if you are unable to beat these runners at 1 mile or 10 km, you will also never beat them at any other distance, even up to 700 km!"

By the way, for you new-comers to the sport, the best $21.95 you can spend to learn the sport, is Dr. Tim Noakes' book "The Lore of Running"

George Parrott

Nick, While you are one of my icons, and your contributions to ultra-running are vast, on the topic of marathon to 50 mile times, I respectfully side with Charls on this one.

I am one of those sub 6 hour 50 mile people, and I ran 5:57 on a so-so day and performance. I had "announced weeks in advance" that I would break 6 hours, but virtually NOBODY thought I had a chance, since I had run "only" about 2:48 at the time for a marathon PR.

However I HAD done lots and LOTS of mileage and ran LOTS of those miles at or near 7:00 mpm and MUCH of the time I was running "tempo runs" of 10-16 miles at sub 6:40 pace.

My fastest 400m ever was a 72.3 seconds and I have NEVER broken 33 for 200m on a track workout. Not then, not ever.

But I was running 120-180 miles a week; I was running 3 "hard" sessions a week, and I was running doubles on speed days including as much as 25 miles in the morning and then doing 4 x 1 mile in the 5:40 range as part of an 8 mile "evening" workout.

I went through the marathon split at 2:56 feeling biomechanically comfortable, but VERY troubled with major blisters. I did a shoe change at about 30 miles and this saved me, for I was able to run the last 5 miles that day UNDER 34:20.

I think, without the blistering I was "ready" to run 5:42-5 territory, and I was certainly ready to run a very good 100km that day, as upon finishing I quickly grabbed a friends bike and rode out several miles on the course to see how my friends were doing....I still had blister problems, but think even that day muscle-wise I was ready to run several more miles.

Later I did improve my marathon PR to 2:41:52.....and ran a 58:10 miler 7 days after that marathon PR.

Daniel Dreyer

I've been away for a while so this response is a little late in coming. I do 2 speed workouts each week. One is an interval workout and the other is what I call a "stride-out" which is usually about 6-8 miles which starts off at a medium pace and has negative splits. I can't say enough about how much these two runs add to my running.

I've also had my own anxiety about doing the runs when the ol' body just didn't feel like going fast. The biggest thing that has helped my is this principle: Don't start off your speed workouts fast! Warm up for a mile or so and then get into it. If you're doing intervals don't do the first ones at top speed or the last ones will kill you. The idea behind doing speed work is to get more and more relaxed as you get faster and faster, NOT more and more uptight. Start off with a speed that feels comfortable but slightly faster than you normally run and don't pick up the speed too fast too soon.

If these tips don't work there are also many other ways to do speed work that can be fun. One of my favorites is fahrtleks (sp?) I do my normal warm-up and then I pretend that my body is a sports car with a 4 speed gearbox. As I'm running down the road I pick a point ahead of me that can be an arbitrary distance (no more than a block) and pick a gear to do that distance in. Before I finish that distance I pick another arbitrary point to get to and another gear to use. These are my gears: 1st gear is a warm-up jog, 2nd gear is a very slow run, 3rd gear is my normal sustainable running speed, and 4th gear is a sprint. I choose the next gear randomly and I jump around through the gearbox and I also can change the distance in which I run any particular gear (this usually depends on how my legs feel).

Mix it up and have fun! Give me a holler if you need more tricks to get rid of those evil speed meanies.


Let me generalize and pigeon-hole ultra runners for sake of simplicity, and this topic. If your goal is "just to finish", then plod on. Plodding will get you to the finish line. If you want a PR or to discover your real potential, I've come to the conclusion that SPEED rules. I base this on having HAD some leg speed, and slowly losing it via age but mostly, I feel, by omission of quality workouts. I enjoyed running as many ultras as possible because of all the reasons we all do them, yet my times were slowing, slowing... In the recent months I am trying to "re-acclimate" myself to quality to regain my leg speed, and the results have surprised me, even through I have NOT (yet) implemented a regime of track workouts (ie 400 m, 800m, or mile repeats.) I feel to run a quality 50 miler, you need to be in shape to run a quality marathon. And if you have the "speed?" to run a 50 mi event in 8 hours; then when you enter a 100 mile/24 hr race, and your 50 mile split is in the 10 hr range, (and all else goes well), you can reach your 100 mi in 24 hours.

Also I am empathetic with the 26-28 hr 100 milers who want to achieve it in 24 hrs, by plodding. In training they plod on their short run, plod even slower on their long runs, and expect to reach 100 miles in a day. I feel the 26-28 hr range runners, are the 24 hr pace runners that had a bad day, and the 30 hr runners that had a good one. ....this is totally my worthless opinion

Finally what Karl K stated below:

"Don't do your long runs at a pace slower than your target pace and then expect to run much faster on race day - training at your target pace will find you running that pace on race day."
I find to be very accurate. Using my weekly10 mile "hard" run as a guide, (yes I know this is not a long run) I've found I can hold that pace for a 50k event, but no-way can I run faster than what I've trained. ...track workouts are on my agenda next, then hills (short ones), and I'll let you know the results.

Pete Petri

Ron Christensen and Blake Wood posted interesting comments on speed work for ultramarathoners.

"Speedwork is sometimes associated with higher injury rates. Coe and Martin in their book "Training Distance Runners" suggest that the best way to avoid injury is to do fast running year 'round instead of adopting it for "sharpening" near a key race. Suddenly inserting a bunch of fast running into a training program is likely to produce injury with loss of fitness."
Karl points out an interesting bit of info, one that I have been chewing on for a while. I keep trying to get back to doing speed work, but every time I do, I end up feeling like crap within just a few weekly sessions and bagging the idea. So how does one breeak into it so that they can begin incorporating it as a regular part of training?

One idea I have is to try to develope a base through the use of a strengthening period. I started doing hill repeats about a month ago, and plan on continuing these as a weekly workout for about another 8 weeks.At that point I will change that night to track work to see if the improved strength helps any.

For hill repeats, I set the timer on my Ironman to 1:30, and run a hard pace uphill, marking the spot I get to after the minute & a half. Jog easily down, then repeat, trying to make it back to the original mark for all other repeats. I don't always make that, but do about 5 out of 6. Last night I did 7 of these repeats, and plan on building to 10.

I hope this will be a help in getting back into speed work again, but I have a feeling I also need to learn how to pace myself a little better too. I may have been going a little too "all out."

Brian Pickett

Al Howie brought up (in a round about way) a question that I have had for quite some time. What do most of you do for speedwork? I guess my intervals are probably shorter than what most of you do, but here are the track workouts and the pace that I hold.

As I said, my intervals are probably shorter than what most of you would do, because I also have to keep my speed up for cross country season when it comes back around. All the same, I'm curious what people who do nothing but ultras and marathons do.

Shawn McDonald #2

Brian Pickett inquired about speed training sessions for ultras. I thought I would share a few of the workouts that I have found helpful in preparing for ultras (been running them since '91). For more background info, see the web page:


For strength development, hill repeats are great. Find a hill (either paved or good non-rocky trail) about 1/4-1/2 mile long. Do a few miles easy to warm up, then do running repeats uphill and recovery jog on the downs. Intensity should be moderately high, I try for about the same effort as I put out for a 5-10k race on roads. The first couple of sessions just do a few (about 4-5) repeats of shorter length (.25 mile kind) and then add a repeat or two per session. Doing these once a week for several weeks will improve your strength and your ability to run more efficiently uphill in races. One can also do power walk repeats on steeper hills. These could be on a hill of 10% or more grade with a brisk stride and intensity slightly above your normal long run effort. If doing the running type hill repeats look for a hill of about 3-4% grade. End your hill workout with a few miles of easy running to cooldown.

For more of speed or cruising ability (ultra specific) I like two main kinds of workouts. One is mile intervals. I run these on a road loop that is marked at 1/4 mile increments. These workouts are best done after you have done a few fartlek workouts and hill workouts in previous weeks. That builds up some fitness base for this harder type of session and prepares your muscles and mind to go faster. A standard workout would be 3 by a mile, with a good intensity on the intervals (about your 10k effort intensity), and a recovery between the fast miles of about 1/4 of a mile or 2 minutes. Start and end the session with a couple of easy miles of running. Another good workout of this cruise type is the tempo run. Here, you warmup with 3-4 miles of easy running, do a few strides (about 100-200 meters each) at a good turnover, and then head into a period (maybe start with 12 minute duration, add 2-3 minutes each or every other session) of faster running. The effort during the faster running period should approximate that you would have during a 10 mile race. This is somewhere about where your anaerobic threshold is. For more info, see the book titled "A Competitive Runners Handbook" by Bob Glover. Follow the faster run period with a couple of easy miles to recover and cooldown. The tempo run works to improve your running efficiency and pace at which you go anaerobic.

The speed sessions for ultra are only limited by your imagination. To prevent overtraining I recommend doing only one speed session per week. Be careful when you first add faster running to your training. It will take your body a few weeks to adapt. The types of speed running that shorter distance runners do can be modified by location and your personal tastes and goals to fit with running and racing long distances. The workouts detailed below are not that different than those I have done in the past for medium-long (half marathon-marathon) distance races.

To put some numbers and framework to the above text, three workouts I did of these types in the past year were:




Adjust the paces and duration of the faster running do your own abilities and current fitness level. I usually run my marathon races in the range of 3 hours plus or minus a few minutes.

Jeff Riddle

My basic speedwork pattern goes like this:

All on grass/dirt xc course, other days easy runs, one day rest.

T - 1000 m x 5 at 5k pace with equal rest or 5 x 1 mile at 10K pace with half time rest

TH - Hill repeats x 4 work up to 12, working on form running down and up or 4-12 x 200m at mile pace with full rest

SAT- Fartlek runs 5,4,3,2,1 min fast runs with 1 min easy run on hilly grass/dirt course for 10-25 miles

SU - 6-9 miles at marathon pace

Do this for about 6-8 weeks then shorter speedwork on track.

Shawn McDonald #3

I thought I would explain what I meant by strides. These are short (100-200 meter) segments of running at a brisk but not all out pace. Their use is to work on leg turnover, running "upright" (i.e. good form, proper arm swing), and as a warmup for more intense periods of running work. In that latter case they are a "bridge" between the easy warmup miles which are done at a conversational pace and the intervals, hill repeats, etc. which are at a higher effort level. Usually 4-5 strides will suffice to get you readjusted to the mechanics of running faster, to warm up muscles and joints, and to get mentally ready to roll. The form work type strides are done after the end of a speed workout, typically on a softer type surface (i.e. grass football field). For these, accelerate over the first half of the stride and then maintain speed, staying loose in the upper body and with your back straight up and down. The strides should not leave you overly winded if you are doing them at a brisk (5k pace or a bit faster) pace. Even better is doing strides in front of a coach who knows what to look for in running form and is familiar with your form and running history.

I agree with Kevin about junk miles that some runners put into their training programs. In the coaching that I have done, it is usually the case that runners want to do more than I recommend during their base building periods. The taper period is another time when junk miles can do harm. Instead do just a short run (thus the 7 mile limit that Kevin set for the point penalty, I might draw the line at 40 minutes regardless of your pace) or a short 30-45 minute cross training session. Rest is another tool in your training program, helping you to ramp up to increasing levels of fitness while lessening the chances that you incur a serious injury.