Frank Horwill



  • These articles were first published many year's ago and whilst some are as relevant today as they were when new, many are now mostly of historical interest as modern research and coaching methods have superseded them.

Shaping Up to the 10km Challenge

By Frank Horwill

With the conclusion of the cross country season many athletes turn their thoughts to racing 10,000 metres, mostly on the roads.

There are good reasons for doing this. First, the distance is no problem to a regular male cross country runner and even women’s cross country races are more than half the distance. The other reasons are that the 10km distance is a good speed workout for the marathoner and for many, the most important factor of all, there are some good monetary prizes at stake - especially on the continent!

It was once thought in Britain that if a man could get under 28 minutes for the distance, and a woman sub-33 minutes, then they were the cat’s whiskers. And so they were... 20 years ago! Between 1977 and 1987, 15 British runners ducked under 28 minutes, and five women under 33 minutes.

However, times achieved then are more than a minute behind today’s world-class times for men and women. Now, only one man from Britain appears in the world all-time 10km list and that time was done 10 years ago.

British women 10km runners have done slightly better, as they have two athletes still listed with times run in 1991.

Russian coaches believe an athlete’s 10km time x 4.7 = potential marathon time. Let’s look at those figures in practice:

Ingrid Kristiansen has a 10km time of 30:13.74. If we multiply that by 4.7 we get about two hours 22 minutes. Her best marathon time is 2:21:06. Another example is Carlos Lopes’s 10km time of 27:17.48 x 4.7 = 2:07:36 approx. His best marathon time was 2:07:12.

Generally it should be noted that for some unknown reason women convert their 10km times into potential marathon times far better than men do.

It is suggested that women do not hit ‘the wall’ as badly as men do at the 22-mile mark of a marathon due to some innate fat reserve. The subject has not been fully resolved.

So, what’s involved in a 10km race? On a track, the physiologist A.V. Hill stated it was 90 per cent aerobic and 10 per cent anaerobic. Theoretically this means that, given 10 training sessions, nine are aerobic and one is anaerobic. We have to decide what is meant by those terms.

The hackneyed view of aerobic running is the ability to run with a partner, starting off with a conversation about politics, TV and sex, and returning an hour later still talking about politics, TV and sex. Such a run would be at 99 per cent aerobic pace, more suited to the marathon.

Training at the target 10km speed is a logical thing to do. A woman aiming to run 32 minutes must consider sustained running at that speed which is 76.5 seconds per 400m, while a man aiming for sub-27 minutes must get used to long spells of 64.5 seconds per 400m.

It looks daunting, and it is, but a start has to be made. We can start with repetition 800ms at those speeds and add to them while still maintaining the original rest time. Most runners above average club level can run several laps at that speed and have a cup of tea and bun after each repetition!

Unfortunately, world class 10km running does not permit any recovery after each lap. We have to get used to that by starting small and getting big.

We may opt to run 12 x 800m in 2:24 with 90 seconds recovery (30 min 10km pace). When we can handle that, the next stage is 10 x 1km in three minutes with the same recovery as for the 800ms. By this means we may eventually be able to do 3 x 3,200m in 9:36 still with only 90 seconds rest. When we reach that point we can truly say we are training logically to improve our 10km time.

There is an age-old coaching axiom which states: ‘in order to run a good 10km, you must first run a good 5km.’ How true is this?

Dave Bedford ran 27:30.80 10km in 1973, a UK record. In 1972 he ran 13:17.21 for 5km. The first is 66 seconds per 400m, the second is 63.5 per 400m, just 2½ seconds per 400m difference.

Bedford’s 5km time doubled plus 56 seconds equals his 10km time. Liz McColgan ran 15:01.08 for 5km, and 30:57.07 for 10km. Her 5km time doubled plus 55 seconds equals her 10km time. We can safely say that an athlete’s 5km time doubled plus 60 seconds equals the potential 10km time. This means two things. Firstly, the athlete’s 5km time must be reduced to its minimum before 10km racing is done. Second, continue to train at 5km speeds while racing 10km.

Training at 72 seconds per 400m (15 mins / 5km) is certainly going to make running 10km in 31 minutes (74 secs / 400m) feel easier. In fact, one of my athletes with a time of 13:11 for 5km who ran a 10km race, said afterwards: "I felt as though I was jogging for half of the race."

Need it be said that a good 5km time is also dependent on a speedy 3km time and that distance is boosted by a good 1,500 metres performance?

In terms of aerobic allocation the 5km distance is 80 per cent aerobic, the 3km 60 per cent and the 1,500m 50 per cent. The last distance will provide the 10 per cent anaerobic allocation for the 10km.

Based on Hill’s findings we can now plan out 10 training sessions over a period of 14 days which will be specific training for the 10km:

Day  Training Session
Day 1 94 percent aerobic – 13 miles at 16 sec/mile slower than for your best 10km time per mile
Day 2 Recovery run for 35 minutes
Day 3 90 per cent aerobicrun 3 x 2 miles at target 10km speed with 90 secs rest
Day 4 Recovery run – 35 minutes
Day 5 60 per cent aerobic – 5 x 800 at 3km target pace with 90 secs rest
Day 6 Rest
Day 7 80 per cent aerobic run 4 x 1 miles at target 5km speed with 90 secs rest
Day 8 Recovery run 35 minutes
Day 9 50 per cent anaerobic – 8 x 400m at target 1,500m speed with 90 secs rest
Day 10 35 mins recovery run
Day 11 Start with day one again

The more serious athletes will consider morning and evening runs of 35 minutes duration each on non-repetition days. Also, they should plan that once a week such a run is done over the hilliest possible terrain.

In 1976, Fox and Matthews threw a spanner in the works of Hill’s deliberations on the 10km. They spoke in a new language of ‘energy pathways’ and chemical symbols. The late chief national coach, Geoff Dyson, and the late national long distance coach, Ron Holman, both felt training the ‘energy pathways’ would revolutionise training methods and result in unbelievable world records.

It is by no means confirmed that the recent spate of world records is due to training ‘the energy pathways’. Fox and Matthews stated that the 10km distance used the following energy pathways: 80 per cent aerobic, 15 per cent LA-02 work

Specimen sessions which use these pathways are:

  • 02 = 6 x 1200m with half the time of the repetition as rest, for example if the rep was done in 3:12, the recovery would be 96 seconds.
  • LA-02 = 8x800 in sets of 2x800m. The recovery was the same time as the time of the rep. For example, if the rep was 2:08, the rest would be the same. After each set the rest would be double the rep time.
  • ATP-PC-LA =16x400 in sets of 4x400m. The recovery was twice the time taken to run the rep. For example, if the 400ms were in 60 seconds the rest would be 120 seconds. After each set the rest was doubled. For example, four minutes in this case.

It is suggested that 02 sessions are done at 10km and 5km speeds. The LA-O2 work is done at 3km speed. And the ATP-PC-LA sessions are done at 1,500m speed. In practical terms this is seven sessions out of 10 allocated to aerobic running, two to LA-O2 and one to ATP-PC-LA.

But, supposing you are like one national coach 20 years ago who said: "Who do these physiologists think they are, telling us how to coach athletes?" One great runner, Ron Clarke (Australia) who broke world records from 3km to 10km, was aligned with that view. After a two-year lay off, Clarke returned to running and shunned repetition work. He confined his plan to running fast for long spells twice a day, his cruising speed being 5 mins / mile pace.

In fact, his training was virtually a race a day! One day he would run three miles flat out, another day six miles at maximum speed. He also raced in local 800m and 1,500 metres races for speed. He was a big weight training enthusiast and did it every other day. He believed in the painful process of racing himself fit, and on one European tour ran 12x5km and 12x3km races in six months as well as half a dozen six mile and 10km events. An Olympic gold medal eluded him. However, there was method in his madness and his fondness for shorter distance races paid off: his 10km world record of 27:39.4 set in 1965 would be welcomed by many British runners today!

To sum up, the weekly key sessions for 10km success are:

  • One run twice the distance of 10km at 16 secs per mile slower than for your best 10km per mile
  • A session at your target 10km time, lasting 10km in total
  • A session at your target 5km time, lasting 6km in total
  • Running up and down a hill where the ascents total 5km, starting at 2km.

Dr Peter Snell, University of Texas, asked 10 runners with 10km times between 34-42 minutes to log up 50 miles of steady running weekly for six weeks. Then they were divided into two groups for 10 weeks. One group ran two lactate threshold runs weekly for 29 minutes (12 secs per mile slower than 10km pace). The other group did two repetition sessions weekly at either 200 or 400 metres which ranged from 10km speed to 3km speed. The repetitions totalled three miles per session. Both groups continued to log a total of 50 miles steady running.

At the end of the 16 weeks both groups were tested at distances from 800 metres to 10km. The repetition training group improved 800m times by an average of 11.2 secs. Threshold trainers improved on average 6.6 seconds. In the 10km test, the threshold trainers improved by 1.1 minutes, the repetition trainers improved by 2.1 minutes.


Because the repetition trainers trained faster than the threshold runners they developed better economy, coordination and comfort while running fast. The repetition trainers were doing more specific work for the 10km. The threshold trainers simply were not accustoming themselves to 10km speed.

In line with this, Yobes Ondieki when training to break the world 10km record, did 10 x 1km at a pace just inside the world record with short recoveries. He said after breaking the record: "My world record (26:58.38) actually felt easier than my tough interval workouts."

Snell’s rep runners spent just 31 minutes per week carrying out their actual interval training, while the threshold runners spent 58 minutes per week This means that the rep runners doubled their 10km performance while carrying out half as much quality training!