How fast should a long run be?

In March 2009, when I was weighing up the possible plans for summer training in preparation for a half-marathon in September of that year, I had explored the Furman program [1].  It is an attractive program for a runner with limited time available for training because it comprises three running sessions per week: an interval session, a tempo run and a long run, together with a flexible amount of cross-training.  At that time I wrote: ‘The thought of a weekly interval session and a weekly tempo session is not too worrying.  Two such sessions are typical of many moderate or high intensity programs.  However, the Furman session that most concerns me is the long run.  The Furman program emphasizes the importance of achieving the designed paces for all sessions.  In the 18 week half marathon program, the pace for the majority of the long runs is Half Marathon Pace + 12 seconds per Km or faster.’   

At that time, I decided on a program that was more closely related to a Lydiard approach consisting of a mixture of low, mid and upper aerobic runs, though far less in total volume than the 100 miles per week that is widely regarded as the defining feature of Lydiard’s approach to base building.   As things turned out, my training was seriously disrupted by a prolonged and debilitating illness in mid-summer and I ran the half marathon in September one the basis of very limited preparation.  This year, a half marathon in September will be my main target race, and I am again faced with planning the details of my program.

Unfortunately I have already had some health problems this year, and am struggling to recover the level of fitness I had achieved in January.  Nonetheless, I still have about 16 weeks to prepare for a race in September. My primary goal is to build up my aerobic capacity, though I will also continue to work on rebuilding some of the leg strength that I have lost in the past decade.   For achieving aerobic fitness with a limited time budget, tempo runs are probably the most efficient type of session, but to minimize the risk of recurrence of the illness that has hampered me, I can only afford to indulge in tempo runs sparingly.  So the weekly long run will be a crucial part of the program.  But this brings me back to the question that I faced when had looked at the Furman program last year: how fast should the long runs be?

The answer depends at least partly on what other training one is doing, and on one’s specific goals.  In preparing for a half marathon the crucial thing is establishing the ability to sustain a pace near the upper end of the aerobic zone for 21 Km.  At first sight, the Furman long run at HMP + 12 seconds or faster seems the obvious answer.  Observing the progress of various runners who have recorded their experiences with Furman on their blogs has confirmed the evidence from scientific studies indicating that the program works – a runner who copes with the program has a good chance of achieving the target race time. However, most runners report that it is tough.  Despite the fact that there are only three specified runs per week, I suspect that in many instances, the runner is on the dangerous far-side of the sweet spot – and verging towards the territory where the risks of injury and ill health rise disproportionately.

 Looking towards Africa

In an era when Africans hold virtually every world record for distances ranging from 1500m to the marathon, the most obvious starting point in addressing the question: ‘how fast should a long run be?’ is to examine the training patterns of Africans.  In fact there is no single African training pattern, and in any case several factors other than current training are likely to contribute to the prodigious success of African distance runners.   However, whatever the contributions of geography, culture or genes, training almost certainly plays a substantial role, and it is perhaps the one feature that might be adapted by any runner.  So it is worthwhile examining the approaches to the long run adopted by contemporary Africans to see if there are any essential features can be identified.  

I have chosen fours runners: two youngsters from Kenya who have made spectacular debuts in international marathons within the past few years, but who have quite different approaches to training, and two ‘veterans’ who dominated international 5000m and 10000m  track events at the end the final decade of the last century, and have between them, held the world marathon record for much of the first decade of this century. 

The two youngsters are Samuel Wanjiru, who won the gold medal in Beijing in 2008 shortly before his 22nd  birthday setting a new Olympic record of 2:06:32; and Robert Kiprono Cheruiyot, 21 year-old winner of the tough Boston marathon in a record time of 2:05:52 in 2010.  The two ‘veterans’ are Paul Tergat from Kenya and Haile Gebrselassie from Ethiopia, who both reached their peak on the track around the turn of the century, before moving to the marathon.   One of the highlights of the Sydney Olympics in 2000 was the titanic battle in the 10,000m in which Gebrselassie narrowly defeated Tergat.  Tergat ran his debut marathon in London in 2001 and in 2003 became the first person to run a marathon in under 2:05 when he established a world record of 2:04:55 in Berlin.  Gebrselassie made his marathon debut in London in 2002, and in 2007 in Berlin took 29 seconds off Tergat’s world record, before going on to become the first person to break 2:04 with a time of 2:03:59, a year later, again in Berlin   

Samuel Wajiru

Japanese journalist, Akio Harada of the Asahi Newspaper, obtained a fascinating insight into the training of Samuel Wanjiru during an interview before the Fukuoka marathon in 2008 [2].  Wanjiru had established a new course record of 2:06:39 in his international marathon debut in Fukuoka the year before.  Harada’s question to Wanjiru about Catherine Ndereba failing to win gold in Beijing because she had not realized until too late that Constantina Tomescu was ahead of her provided an interesting insight into Wanjiru’s mental approach to racing.  Wanjiru had not seen the women’s race but his comment was ‘but I’m not the type of guy who runs behind someone else.’   In contrast, his approach to training was perhaps best encapsulated in the remark; ‘Sundays are off, and if it rains I also take the day off from training. If you train too hard in the summer it’s bad for your body.’

He described his training in the months leading up to the Beijing Olympics in August 2008.  He did long cross-country runs of 38 Km; somewhat faster runs of 30Km on flat ground; and speed work, such as 10x400m or 3x3000m.  In addition, about half of his training days were described as easy days.  The 38Km long runs were at a pace of 4:30 per Km (7:12 per mile).  This pace is 50% slower than his marathon pace.  For a 3 hour marathon runner, the equivalent pace would be 10:17 per mile.  It should be noted that these runs were at altitude and also cross-country.  As an aside, I was interested in his comment:  ‘I never do any strength training. You can get enough strength from running on cross-country-type courses.’   Nonetheless, his long run pace suggests that many of the benefits of the long run can be obtained by running at a very gentle pace; a pace that is ‘not bad for your body’.    His 30Km ‘pace runs’ were run at a pace only 5% slower than his marathon pace, and hence might be regarded as tempo runs.  

Thus, as in the Furman program, Wanjiru’s training included speed sessions, tempo and long runs, but the really striking differences are that his long runs were much slower than the Furman marathon program would recommend, and his total volume was higher.  In particular he added substantial volume on his easy days, some of it apparently in the form of progressive runs. He told Harada:  ‘Other than that, I run about 15 km at 7:00 or 8:00 in the morning before I have breakfast. I’d say it’s like jogging, or a buildup to something like a pace run. I start around 4 minutes per km, drop it down to 3:30 in the middle, then end up down at about 3 minutes 5 seconds. I always want to finish thinking, “Aaaaah, that felt great. I had a good run.”’

Robert Kiprono Cheruiyot

Robert Kiprono Cheruiyot made his international debut at the Commerzbank Marathon in Frankfurt in 2008, winning in the new course record time of 2:07:21.  A year later, Larry Eder published details Cheruiyot’s training leading up to the 2009 Frankfurt marathon on RunBlog [3].  At that time, Cheruiyot was running around 280 Km per week in training, in contrast to Wanjiru’s total volume of around 140 Km per week.  Cheruiyot also included a mix of long runs (typically 40Km, midweek); a somewhat faster paced 30Km on the weekend; and a weekly ‘speed’ session.  However his typical speed session was 12x1000m at 3:00 min/Km, which is slightly slower than his marathon pace and hence is likely to be comfortably below the anaerobic zone.  Eder did not report the pace of the 40Km long runs, but the schedule makes it clear that the majority of Cheriuioyt’s training volume was provided by 15-16Km runs at a pace about 1 ½ minutes per mile slower than his marathon pace.  Thus the overall picture is of a program with very high volume but only moderate intensity.   

Paul Tergat

However not all of the Kenyans place quite so much emphasis on avoidance of excessive stress.  Among those with a legendary ability to tolerate hard training is Paul Tergat.  The biography ‘Paul Tergat: Running to the Limit’ by Jurg Wirz [4], emphasizes Tergat’s commitment to developing mental toughness.  The training programs described in the book appear relentless in their demands.  For example during his preparation for the 2002 London marathon, the  total volume weekly volume was  well over 200Km per week including 7 quite demanding morning sessions and 6 easier afternoon sessions.  The 7 morning sessions typically included one or two sessions of speed work, a long run and three or four sessions of 70- 75 min which are described as ‘always medium to fast’.  The long run is described as ‘not slow at all’.  The easy afternoon sessions were typically 1 hour at a slow to medium pace.     

Nonetheless it is noteworthy that unlike many other Kenyan distance runners, Tergat did most of his training at a relatively low altitude of around 1700-1800m, near his home in Ngong on the outskirts of Nairobi.  In 2001 he had trained for about two months at an altitude of 2300m in the Kaptagat Forest at Eldoret, but he found this very exhausting.  He reports that ‘by evening I was very, very tired’.  Thus, despite the apparent Herculean character of his training schedules, even Tergat was aware of the need to avoid getting too tired.

It is also noteworthy that following the exhausting high altitude training in 2001, he was second in the London marathon in a time of 2:08.  While this is a very creditable debut marathon, when the Morrocan Abdelkader El Mouaziz  had applied the pressure at 35 Km, Tergat could not stay with him.  In 2002, Tergat was again second, ten seconds behind Khalid Kannouchi’s new world record time of 2:05:38.  In 2003, Tergat reached the high point in his marathon career with his world record of 2:04:55 set in Berlin. Had he perhaps trained just a little too hard in 2001?  And, since 2003, he has not achieved the same heights again.  His most recent, and perhaps final, marathon victory was in the 2009 Lake  Biwa Marathon in Japan, which he won in 2:10:22.  It would be churlish to criticize the achievements of Tergat’s remarkable running career, a career than has earned him  a place in the Pantheon of running greats, but I can’t help wondering whether or not he might have sustained his marathon performance at the very top level  for a few years longer if he had adopted a slightly more relaxed attitude to training. 

Haile Gebrselassie

While Tergat has earned a place in the Pantheon, in my opinion the man who is entitled to be described as the greatest distance runner of all time is Tergat’s friend and rival, Haile Gebrselassie.  Not only is he the current world marathon record holder, he has established a total of 27 world records over distances from 2000m to the marathon during his career, and in addition, has run some superb competitive races, of which the 10,000m final in Sydney stands out.  However, as an old timer hanging onto a shred of nostalgia for the days of amateurism, it is with wry amusement that I am forced to go to the website of one of Gebrselassie’s commercial sponsors, Powerbar, for details of his training. 

The example of Gebrselassie’s training provided on the Powerbar website [5]  demonstrates that his a weekly program includes speed work, a tempo run and a longish run together with a fartlek session and easy runs giving a total weekly training volume of 190 Km (120 miles) .  His speed work consist of 6x2km @ 2:50 /Km which is slower than his best 10Km pace, with 3 min recovery; and also 10x100m sprints within two of his easy sessions.   In his tempo session he covers 15Km in 48 min.  (i.e. 3:12 /Km) which is about 7% slower than his marathon pace.  His longish run is 25Km in 90 minutes  (i.e.  3:36 /Km or 5:45 /mile; for a three hour marathon runner that would correspond to 5:14 /Km or 8:22 per mile).

The easy sessions include a 20 Km run at 4 min per Km and two 18Km runs at a similar speed, in addition to five easier runs, such as 10Km in 43 minutes (4:18 per Km, which is 46% slower than his marathon pace; for a three hour marathon runner, that would correspond to around 6:15 per Km or 10 min/mile). 

In summary, Gebrselassie’s program is fairly high in both volume and intensity, but even the higher intensity runs are not very intense.   His 25 Km long runs are shorter than the corresponding 38 Km runs of Samuel Wanjiru but somewhat faster.  The pace of Gebrselassie’s 25km runs is about a minute per mile slower than his marathon pace.  Some of his easy runs appear very easy – such as 10Km in 43 minutes which is a pace equivalent to 10 min/mile for a 3 hour marathon runner.  It should however be noted that much of his training is done at an altitude of around 2700m. 

In an article about her experiences arising from Gebrselassie’s invitation to her to visit Ethiopia, Canadian 1500m runner Hilary Stellingwerff reports that Gebrselassie placed great emphasis on mental preparation; on the need to deal first with one’s inner demons.   “First you must win yourself. Of course you must train hard, but after that, everything is easy.’ [6]  She describes her humbling experiences as she joined in with the local athletes running a 16Km loop in the mountains not far from Addis Ababa and concludes “Athletics comes number one for African athletes; there is nothing else. They train very hard and sleep the rest of the day to recover.’  However she also points out that they respect the body.    She writes; ‘Finally, on all my recovery runs, the Ethiopian athletes stressed the importance of running on soft ground in the forest to make sure you go slow enough to really recover. They don’t worry too much about their pace, but instead about “getting good oxygen” from the trees and “soft ground” for the body.’


As might be expected, there is no single answer to the question of how Africans train, but several clear themes emerge.  Elite Kenyan and Ethiopian distance runners adopt a program comprising speed work, ‘pace’ or tempo sessions, long runs and easy runs.  The speed of the long runs is at least 1 min per mile slower than marathon pace and often slower.   The easy runs appear very easy by European standards.  Strong determination is balanced by an over-arching spirit of respect for the body.

Overall, the spirit that can be distilled from these accounts of the training of elite Kenyan and Ethiopian distance runners is perhaps a little nearer to the spirit of Ed Whitlock (discussed in my post on 22 April) who trains 3 hours per day, mostly at very low intensity, than it is to the remorseless pressure of the high intensity Furman program, despite its low volume.   I still consider the Furman program has much to recommend it, but I am even more certain that my initial concern, expressed last March, that Furman long run paces are too demanding, was well founded.   In the months ahead, I will aim for long run paces that are at least a minute per mile slower than a speculative estimate of my marathon pace.

London Marathon 2010

As a footnote to my posting on the Sweet Spot on 22 April, Marie and her husband did cross the line together on Sunday in the 2010 London marathon.  The truly inspiring achievement was Marie’s determination to celebrate the experience and to get her medal. 

Another inspiring achievement was the 2:47:34 recorded by Rick Bowker, the ‘No limits desperado’ of Rick’s Running.  He had prepared tenaciously following Marius Bakken’s 100 day program.  Finishing in 376th place overall (and 30th  in the 45-49 category) while taking around 4 minutes off his PB was a great achievement.   Well done, Rick, and thanks for all your comments on this site over the past year.


[1] Bill Pierce, Scott Murr and Ray Moss (2007) ‘Runner’s World Run Less, Run Faster: Become a Faster, Stronger Runner with the Revolutionary First Training Program’

[2] Akio Harada (2008)  ‘Samuel Wanjiru shares the secret of training to win.’, published in the program for the 2008 Fukuoka marathon; translated from Japanese by Brett Larner with editorial assistance from Mika Tokairin, and posted on

[3] Larry Eder (2009)  ‘Robert Kiprono Cheruiyot: training log, from Victah Sailer, Notes by Larry Eder’.  Posted  23 Oct 2009 at

[4] Jurg Wirz (2005) ‘Paul Tergat: Running to the Limit’


[6] Hilary Stellingwerff (2009)  ‘Living the High Life.’  Canadian Runner, Issue 2.4


17 Responses to “How fast should a long run be?”

  1. rICK Says:


  2. RICK Says:

    con- over a medium distance the following week, alternating.

  3. SPR Says:

    Interesting blog.

    Is there any info on the longest run HG does in prep for the marathon?

  4. Westley Says:

    Canute, excellent article and a great synthesis of material from a range of sources that provides a coherent analysis. It will give me much food for thought as I devise an amended Pfitzinger & Douglas up to 70 miles schedule for my next training cycle. Thanks again.

  5. canute1 Says:

    SPR, Thanks for your comment. With regard to Haile’s long runs when preparing for a marathon, last year when he was preparing for the Berlin marathon, about 4 weeks before the event he posted on twitter that he had just done a 3 hour run , at an altitude of 3000m ( @HaileGebr, 27th August 2009). Since he has stated previously that he never runs any slower than 4:30 per Km, that 3 hour run must have been at least 40Km.

  6. canute1 Says:

    I agree that alternating slow and faster long runs is worth considering . However, in view of the evidence from the elite Africans that a large amount of relatively slow running combined with tempo runs and some speed work, is effective, I am inclined to keep the long runs slow, but in most weeks I will also include a moderate length (12-15Km) progressive run, during which I steadily increase the pace up to half marathon pace – similar to Wanjiru’s progressive runs. My experience is that this type of progressive run is exhiliarating but less stressful than a run of similar length at HMP+12 sec. In the next few months I do not plan to do interval sessions, but following Lydiard (and Haile Gebreselassie), I will continue to include some ‘wind sprints’ at the end of easy runs to try to maintain some speed in my legs.

  7. Ewen Says:

    Thanks Canute for another most interesting post. I’m very pleased to hear of Marie’s success at London.

    Just briefly, I like your idea of the Wanjiru ‘progressive’ run. That also gives one the option of switching the run to ‘easy’ if the body if having an off day. After London, I wonder a bit if Wanjiru has ‘used up’ the base of mileage he ran prior to his current training methods. Or if the 140k/week of relatively high intensity running is less reliable for consistent marathon performance than Robert Kiprono Cheruiyot’s high mileage/low intensity method.

    Out of interest, another amateur runner who has been using the Furman method in recent months is ‘Luckylegs’ . She has run not much over 30 minutes for 5k and 2:22 for the half marathon at age 80. Her cross-training is different though – using either easy running drills, strides, very short hill sprints, or brisk walks on the ‘non-running’ days. She also has found the suggested pace of the long run difficult to achieve.


  8. Ewen Says:

    Ah, yes. I was going to say expect a box of Powerbars in the mail from Geb’s sponsor!

  9. SPR Says:

    Thanks for that Canute, seems they all go quite long. Tadese’s training for a Half marathon would be interesting to see.

    Interesting about Wanjiru and XC replacing weight/strength training. Pete Pfitzinger recommends rithe doing weights or hill running is training This kind of backs up Wanjiru, he is strength training via the XC runs. I go for the weight training option but XC race as much as I can during the winter.

    Started Weight training again in Jan (first time in about 18 months), and I can feel the benefit in my running (Feb & Mar were poor running month due to an ankle injury during an XC race). Basically my HR stats are ahead of where they should be due to just running, and I feel stronger in training runs.

    Hope your training and racing is going well.

  10. canute1 Says:

    Ewen, It is difficult to draw conclusions from Wanjiru’s performance in London on Sunday. This was his first marathon failure after five spectacular performances including victories in Fukuoka (2007), Beijing (2008) London (2009) , Chicago (2009). However his training for London 2010 suffered several disruptions. According to an article posted on on 24the April he suffered a back injury and spent two weeks undergong treatment in January ( To test his recovery he ran the Rock and Roll half marathon in NewOrleans, coming second to Martin Lel. After his return to his home in Nyahururu he suffered bruised knees in a fall during the melee at the start of a crowded 40km training run, and then his speed training was disrupted when heavy rains washed out the Nyahururu Stadium forcing him to transfer the speed sessions to a roadway. There were strong hints in an interview reported by Holly Hamilton of Sportsbeat on the Friday before the London Marathon that he was not fully confident regarding Sunday’s event. He admitted he was a little unsure about how it would go this time. He did not sound quite like the confident young man interviewed by Akio Harada in Fukuoka in 2008. However, Sunday in London was probably an important formative experience for him. Both Haile G and Paul Tergat have emerged stronger after disappointing performances during their careers. I am a little disappointed he actually dropped out on Sunday, though I do not know whether or not his back problem flared up again. Running sub-3 min Kms when your back is aching and you know you are beaten is a very tall order. I was full of admiration for Abel Kirui as he kept going despite falling back to 5th place after Kebede crushed him on Sunday. I am sure we have not seen the end of the great performances from both Wanjiru and Kirui.

  11. canute1 Says:

    As far as I know Tadese runs moderate volume but mainly at an even higher intensity than Wanjiru. He will probably be the ‘youngster’ to watch in the marathon in London 2012 – though he is in fact older than either Wanjiru or Cheruiyot and his transition to the marathon has been hesitant so far. His 2:12:03 in 7th place in London on Sunday does not reflect his potential. While there was some reason to hope that his HM in Lisbon 5 weeks ago might have fired him up for London, it is actually very hard to maintain that peak fitness for two long races 5 weeks apart. (Ewen might point out that it is especially difficult when fitness is based on a high intensity, moderate volume program)

  12. Ewen Says:

    Yes, I agree with that point about maintaining a racing peak off high intensity/moderate volume. Also with Tadese, perhaps he underestimated the recovery needed from a WR half marathon effort?

    I think Wanjiru did the right thing by dropping out of London. I read that invited athletes at London don’t receive appearance money if they don’t finish, hence the jogging finishes of Tadese and Deena Kastor. Although I would have thought both don’t need the money.

  13. Dave from Running Tips Says:

    I know that training can be analyzed and over-analyzed, but the lifestyle that some of these runners live is also a big factor. How much do they rest everyday? Do they train with other runners near their ability? What is their internal motivation? These things are harder to measure.

  14. canute1 Says:

    Dave, Thanks for your comment. I agree that life-style is important. In particular, a large amount of sleep is crucial to fully mobilize the anabolic effects of growth hormone, which released more copiously during sleep.

  15. syl3 Says:

    Why do runners still use altitude training? I thought it was discredited in aerobic sports like cycling and biathlon, now everyone says live high train low.

    • canute1 Says:

      Syl3, Thanks for your comment.
      I suspect that one of the reasons that high altitude training has continued to be of interest to distance runners is that most of the world’s best distance runners train at high altitude, whereas this is less true in cycling and other endurance sports. Whether the fact that most top distance runners train at altitude is due to the effects altitude training in itself, or due to some associated factor (eg running to and from school in childhood; cultural effects on motivation to train; genes or any of many other possibilities) remains a debate. I suspect that few Kenyan children in the mountain villages have the luxury of riding a bicycle to and from school, so we cannot easily test whether the combination of factors that characterize the leading Kenyan distance runners could produce similar benefits in cycling.

      There is at least some physiological rationale for altitude training, especially the greater stimulus to produce hemoglobin. My own view is that high volume, low intensity training at altitude is probably an effective way to increase oxygen carrying capacity of the blood, and probably also to increase oxidative capacity of mitochondria and density of capillaries – all potentially beneficial for running fitness. However I believe that the risk of placing too much stress on the body are substantial unless one has spent one’s childhood at altitude. Therefore for individuals who spent their childhood at low altitude, I think it makes sense to live high and train low. Living high will provide stimulus to hemoglobin production without producing too much stress on the body.

      Overall, I do not think there is any one answer as to how to train, though whatever training program one follows, it is probably essential to achieve a good balance between working hard and allowing the body to recover

  16. veepizsupport Says:, Our online African Village. We would highly appreciate your blogging on our site. You can sign in by Open ID(use facebook, yahoomail, hotmail, etc to sign in) or register with us.

    NB: We are the premier site in the world with A University Research Tool.You can download termpapers, notes, exams and research paers submitted by African students. This feature is completely free.
    Secondly, we have have a lot more features than Facebook(for example, a better chat system, games, chatrooms, university research tool, Open ID sign in, an invite tool and a lot more)
    Furthermore, we are proudly East African. Made in East Africa, coded by East Africans.
    Lastly, we are not trying to copy Facebook but we are trying to be better since it is the benchmark

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: