Jim2's Running Page

Ultimate Speed Workout

About Me
Favorite Links
Contact Me


The subject of speedwork for marathon training has been discussed many times on the RWOL Marathons forum. I think the consensus of most experienced marathoners is that a balanced training program that includes a mix of lactate threshold (LT), hill training, VO2max, anaerobic, and marathon pace (MP) sessions is the best way to improve and optimize marathon performance. Most also agree that the emphasis in such a program should be on LT and hill work, although a growing number also advocate extensive MP training. Except for novice programs, most "cookbook" marathon training programs include all of the above in one form or another and to one extent or another. In fact, most training programs for distances shorter than the marathon also include the same "speedwork" ingredients, except for MP runs, but the mix of them changes for the shorter distances.

However, I believe there is a specific "speed workout" that is superior to all of the above. It provides more training benefit, with little or no additional risk of injury, than the "standard" speed workouts. Run correctly, it is a "balanced" workout within itself that includes all of the speedwork elements, except for MP running. And it is much more fun than any of the other speed training workouts.

So, what is this "ultimate" speed workout? It's 10k racing. Not 5k racing. Not half marathon racing. But, specifically, 10k racing.

A well planned running regimen will include a few races within a training cycle or "season" that are intended primarily for training purposes while enroute to the target race. (When I sit down with a calendar to prepare a training program, the very first thing I write in is the races that I intend to run during the 14-18 week training cycle.) Also, a marathoner's well planned long range regimen will include "off season" racing that is mostly intended for training and development purposes. These races are the hardest and "best quality" speedwork that most runners can do. And I believe that the 10k race distance best serves this training purpose, especially for a marathoner.

Why is 10k racing so beneficial for marathon training? Primarily because it is the most challenging LT workout that most runners can do. And, from a "speed" perspective, LT improvement is the most critical training element to running faster marathons.

A 10k race is essentially a LT race, but a very intensive one. It requires that you run for 40-60 minutes at a pace that begins at and quickly reaches somewhat above LT. (LT pace is generally considered to be the pace that you can sustain for about an hour, which is approximately 10-15k pace for most runners.) It is longer than all but the longest tempo run and is run at an average pace that is a little faster than most runners' tempo pace. I view it as essentially a "super tempo" workout.

A 10k race also provides more than just superior LT training. The last 1/3 of a 10k race is run well above LT. During this phase of the race, the racer reaches VO2max level. Thus, a 10k race includes a reasonably good VO2max workout.

Many 10k race courses include at least some hills. Thus, the 10k racer usually gets some degree of hard hill running, which is important for strength training and running economy.

Finally, the final push to the finish line puts you well into the anaerobic range. It challenges your anaerobic capacity to the max.

Thus, the typical, well paced 10k race includes a combination of LT, hill work, VO2max and anaerobic running with the emphasis on LT and, depending on the course, hillwork.....and that is exactly the mix that a marathoner should be looking for in training. The race keeps most runners on the upper segment of the blood lactate curve (above the lactate turnpoint) for a longer distance and period of time than most other commonly available race distances.

A 5k doesn't provide nearly the same training benefit of a 10k. Primarily because it is half the distance. Thus, you get half the workout of a 10k. True, it is run at a faster pace than a 10k....about 15 sec/mile faster. However, that isn't important to a marathoner. Distance and duration while under intense stress is more important. Bottom line....a 5k race falls far short of the training benefit of a 10k race. Unfortunately for marathoners, unlike yesteryear the 5k is the most popular race distance today.....more on that later.

OK. How about 15k's, 10-milers and half marathons? Aren't they just as good or even better than the 10k for training purposes? After all, they are better suited to "practicing" MP, if that's your training goal. Raced all out, they also get the racer to and above LT for many miles and up to VO2max at the end. Plus, they stress you for a longer time and distance than a 10k. Well, perhaps they are as good as or even better than a 10k in those respects. But there are three disadvantages to them as compared to 10k's:

(1) They are fewer and farther between. Thus, there is less opportunity to use them as training races.

(2) They require more taper in order to race them most effectively. That might mean just another day or two from a training program, as compared to "tapering" for a 10k, but in a marathon program every day counts.

(3) They require more recovery than a 10k. Again, that bites into quality training time.

On balance, I consider the 10k to be the most efficient training race distance. And I think that racing is the most beneficial form of "speed training". That combination makes the 10k the ideal or "ultimate" speed training run in my mind.

Now, what is the downside to relying on 10k racing for speedwork?

One obvious limitation is that you can't run them as frequently as other standard forms of speedwork. You certainly wouldn't consider running one every week while marathon training, as many runners schedule a somewhat shorter and/or slower tempo run every week.

Another consideration is that they beat you up more and require longer recovery than a tempo run or an interval session. It might be just a day or two longer recovery. But, again, every day culled from a training program is a training day lost. And we already lose enough training days in a typical training cycle for other reasons.

The costs of racing in a marathon training program have to be balanced against the gains. The need to spend a day or two tapering and recovering has to be weighed against the training value of racing. Thus, you shouldn't race too often while marathon training, but you also shouldn't not race at all. I think that 4-5 races in a 16-20 week marathon training cycle, or about once a month, is a good goal. And, except for one half marathon late in the program, I suggest making as many as possible 10k races.

The 10k race can also benefit marathoners when not in a specific marathon training cycle. The LT gains are just as meaningful during the "off season" and carry forward into the next marathon training cycle. In fact, I believe that the gains can be even greater during the off season because the runner can race more frequently, race when fresher since s/he is running less mileage, and do more frequent and intense speedwork between races. That is one reason why the macro-cycle plan that I always followed was based on alternating marathon and 10k seasons.....a spring 10k season and a fall marathon season separated by brief (2 weeks or so) R&R breaks. I found that the higher mileage of the 5-month marathon season and the more frequent and intense speed training/racing of the 6-month 10k season complemented each other. Each provided a break from the rigors of the other (the hard/easy concept on a macro level) and each fed its strengths into the other. I found it to be an excellent way to stair step to higher levels at both marathon and shorter (5k-half marathon) distances.

One final "problem" with using 10k racing while marathon training or for alternating seasons. Many runners will find it difficult to find enough 10k's to meet my above suggestions. Twenty years ago the 10k was the staple of a runners program. It was the most popular, thus most commonly found, race distance. It dominated the running world. Today, the 5k has supplanted the 10k as a runner's "baseline" distance....at least in the U.S....and 10k's are becoming harder and harder to find. That has resulted in a significant decrease in racing mileage/year for the average runner. Based on my experience over the past 20 years (see my supplemental info at the end of this post for details), I estimate that the typical American runner races approximately 15-20% fewer miles/year as compared to 20 years ago simply due to the substitution of 5k's for 10k's. And I think that has hurt the American marathoner.

I suggest that any runner who wants to improve in the marathon distance would benefit from running just about as many 10k races/year as s/he can find and can work into his/her schedule. An excellent goal would be to run an average of one/month, or about a dozen/year. You might have to seek them out. And you might have to go out of your way to run some of them. But they are worth it. Many runners will have difficulty finding that many 10k's to reasonably work into their schedules. As an alternative, look for races between 8k and 10 miles. I would opt for a 5k only as a last resort, although even it is better than just another 6x800 interval workout. ;-)

If you are serious about improving your marathon performance, then become a 10k "junkie". I think that you will be pleased with the training "high" that you get. :-)


Supplementary Personal Information.

Twenty years ago in the Baltimore area, I had more than one 10k race from which to choose most weeks of the year. I probably could have run 30 or more 10k's in a year if I had wanted to. However, that began to change throughout the last half of the 80's when race directors and sponsors began to sacrifice distance (read "challenge") to boost field size and improve race management economics. Today, a 10k is often hard to find. I think that has hurt the average American marathoner.

The following describes the trend in race distance mix and average race distance, excluding marathons, that I experienced over the 18 years of 1983-2001. I have broken my data down into three 4-5 year time periods that included comparable total numbers of races. (I took a break from serious running and ran no races during 1991-96.) The highlighted the data is most relevant to this discussion.

Race Distance
































Half Mar




Metric Mar








Total Races




Races, excluding Mar




Race Miles, exc Mar




Ave Miles/Race




Note two basic trends:

(1) Although the number of races were about the same in the 1983-86 and 1997-01 periods, the number of 10k's decreased by 23 from 38 of 67 races (57%) to 15 of 65 races (23%). Meanwhile, the number of 5k's increased by 20 from 2 of 67 races (3%) to 22 of 65 races (34%). That's almost a one-for-one replacement of 10k's by 5k's and represents a heck a lot of training race miles lost.

(2) The average number of miles/race, excluding marathons, decreased by 7/10 of a mile between the 1983-86 and 1997-01 periods. On a per race basis, that doesn't sound like much. However, project that over 15-20 races/year, which is what I typically ran each year, and it is a difference of 10-14 miles of high intensity racing. Extrapolated over a longer term of 5 years, it is 50-75 miles of excellent, high intensity training lost. That becomes significant.....it is important to keep the long term, macro view in mind.

My decrease in training race miles was not by design. It reflects the selection of races that were available to me. When I had a choice between a 5k and a 10k, I usually opted for the 10k. What changed was the available choices. Twenty years ago, 5k races were oddities. On the rare occasion that one was available, serious runners sometimes would run it just because it was something different to do....a change of pace, so to speak. Today, however, the 5k is perhaps the most common race distance.

The bottom line for me is that a lot of racing (read "training" and "development") miles were lost in my second running life, as compared to the early years of my first running life. I think that the decrease in 10k racing affected my rate of development in my second running life as compared to my first running life. I ran about the same number of total races (67 vs. 65) and marathons (9 vs. 8) during both periods. However, I made better progress and wound up running stronger marathons the first time around, even though I started the second time from a much better long term base. I think that the difference in year round racing regimen was a contributing factor.