Contents
Pace and splits
| Pace | 4:44/km |
|---|---|
| 5 km | 23:42 |
| 10 km | 47:24 |
| 15 km | 1:11:06 |
| Finish | 1:40:00 |
Your ranges from the Cooper test
This goal matches an estimated result of about 2850 m in the Cooper test, a 12-minute run for maximum distance. If you are slower today, set your own result and the ranges will follow your current fitness.
This is an estimate, not a promise of a finish time. Your coach sets the exact ranges after a test in the app.
Repeat times in each range
The middle of each range at the estimated Cooper result. Use it to check your pace on a watch or a treadmill.
| Range | 400 m | 1 km |
|---|---|---|
| Easy | 2:31 | 6:17 |
| Aerobic | 2:13 | 5:32 |
| Tempo | 2:00 | 5:01 |
| Threshold | 1:51 | 4:38 |
| VO2max | 1:42 | 4:15 |
Mileage stops explaining the result
Look at the middle rows of this table from a 2023 study of 134 recreational male half marathon runners.
| Group (minutes) | Average time | Days a week | Per week | VO2max |
|---|---|---|---|---|
| 108–117 | 1:50:48 | 4.1 | 45.9 km (28.5 miles) | 45.7 |
| 100–107 | 1:42:54 | 4.4 | 54.9 km (34.1 miles) | 48.1 |
| 93–99 | 1:36:00 | 4.4 | 56.6 km (35.2 miles) | 51.9 |
| 73–92 | 1:24:30 | 5.4 | 67.9 km (42.2 miles) | 54.5 |
Between 1:43 and 1:36 the number of days is the same and the volume differs by 1.7 km (about a mile) a week. VO2max, maximal oxygen uptake, differs by 3.8 points instead. The jump from 1:50 to 1:43 is about 9 km a week more, and from 1:36 to 1:24 about 11 km and one more running day. Right where 1:40 sits, volume stands still and the result improves anyway.
This is a cross-sectional study, so it does not prove cause, but it suggests a direction. At about 55 km (34 miles) a week, another 10 km will probably give you less than better use of the ones you already run. The Cooper test equivalent of 1:40 is an estimated 2,850 m, which is the result behind the ranges shown above the article.
What to train once the volume is there

In a 2017 study of recreational marathon runners, speed at the first and second lactate thresholds, the intensities from which blood lactate starts to rise, was linked to the result more strongly than VO2max (correlations of 0.79 and 0.72 against 0.54). That was the marathon and only 82 runners, but a half marathon is run even closer to threshold, so the finding fits here at least as well.
A 1:40 is therefore built on two things. The first is Threshold work in longer repeats. The second is some faster running around 5K pace, which works on the VO2max difference in the table. For a 2:00 goal that is optional. Here it is necessary. Race pace itself, 4:44/km (7:38/mile), lies at the slower end of the Threshold zone, and you have to tolerate it for 100 minutes. Repeats at race pace make sense late in the preparation, as long as you do not pile them onto a week that is already hard.
Take a peak of 55–75 km (34–47 miles) a week as your reference, the same one we give on the main half marathon page. At 1:40 the lower end becomes risky, and the upper end makes sense only if you can handle it without breaks. Judge it by the quality of your Threshold sessions, because the weekly total on your watch will not show it. A few seconds per kilometer decide whether a session builds fitness or only tires you. That is why an IMI coach sets your pace ranges after a Cooper test or a 5K.
In numbers: at about 55 km, Threshold work adds up to roughly 30–40 minutes a week and the faster running to 10–15 minutes. If today you run everything at around 5:30/km (8:51/mile), the first change is slowing your easy runs to about 5:55–6:15/km (9:31–10:04/mile). Without that, the Threshold sessions will not come off.
How rare a result it is
No official statistic covers every half marathon, so the closest reference is a big city race. A peer-reviewed study split the 133,268 finishers of the Vienna half marathon from 2006 to 2018 into quarters by finishing position. These are the approximate average times, our conversion from the published speeds.
| Quarter of the field | Men, average time | Women, average time |
|---|---|---|
| Fastest | about 1:37 | about 1:51 |
| Second | about 1:52 | about 2:06 |
| Third | about 2:03 | about 2:17 |
| Slowest | about 2:23 | about 2:36 |
For a man, 1:40 lies between the averages of the fastest and the second quarter, much closer to the fastest. For a woman it is about eleven minutes faster than the average of the fastest quarter, a result well toward the front of the field. These are averages from one race, and another field can be faster or slower.
Age changes the picture more slowly than you might expect: in a study of more than 300,000 marathon and half marathon runners, average times were nearly the same from 20 to 49 and slowed by 2.6–4.4% per decade between 50 and 69.
1:40 has to be in your legs before the finish
Around 2:00 the round number works like a magnet and the field bunches just under it. Do not count on that here. In the Vienna data the fastest quarter of men was the only group with no finishing kick: they ran the first 5 km at 4:28/km (7:11/mile), the stretch from 15 to 20 km at 4:41/km (7:33/mile), and the last 1.1 km slower still. At this level there is little left to spend by then, so do not count on the finish line to give you a few seconds. The pace has to be there earlier.
The good news is that this group runs evenly. Its speed dropped by under 5% from the first 5 km to km 15–20, against 13% in the slowest quarter. At 10 km your watch should show less than 47:24, and at 10 miles less than 1:16:17. A cushion of 30–60 seconds at 10 km is worth having if it comes from fitness. If you are at 47:40 and working hard, 1:40 is very unlikely: in that study no group, at any level, ran the second half faster on average.
Do not count on the crowd either. Berlin lists pacers at 1:35 and 1:45, but none at 1:40. In Copenhagen the 1:40 pace team starts in the 1:40–1:50 group, while a runner who declares 1:39 lines up one group ahead with the 1:30 and 1:35 teams. The field is strung out at this pace and you will not always have company, so think about which time you declare and practice holding 4:44/km on your own.
You will not run out of fuel, but a gel can help
A 2010 model estimates that a 70 kg (154 lb) person has about 88 g of carbohydrate in the liver and about 310 g in the leg muscles, roughly 1,600 kcal in total. A half marathon at this pace uses clearly less, so you will not run out. The ACSM recommendation, 30–60 g per hour, means 50–100 g over 1:40. The upper part of that range is hard to take in at this effort. One or two gels, the first at around 45 minutes, are there to help you hold the pace at km 17–20.
Try the gel in a training session with repeats at race pace. If your stomach does not tolerate it, go without. Race day is not the place to experiment.
Start time and a short night

Not every half marathon starts at nine in the morning. Singapore's is flagged off at 4:30 am and Mumbai's waves leave from 5:00 am to avoid the heat. Berlin's runners go from 10:05 am, and the Las Vegas half marathon is run at night. Each breaks the routine of a normal morning run in its own way.
A 2022 review of 69 publications covered people who slept 6 hours or less. In the morning, performance mostly held. In the afternoon and evening it dropped consistently, by about 7.5% on average, with a very wide spread between studies. Do not convert that into minutes. The conclusion is simpler: a late start copes worse with a bad night, and a pre-dawn start makes the short night itself the thing to rehearse.
After a poor night, move the goal from 4:44 to 4:48/km (7:43/mile), about 1:41:15. Before an evening race you have the whole day, so rest and meals are in your hands. Plan the last larger meal exactly as you would before an evening training session, because that is the only place you can test it. Before a dawn race, test the early breakfast on a long run that starts at the same hour. In strong wind and without a group, 4:46–4:48/km is also more sensible than a lonely fight for 4:44.
When 1:40 is realistic: a 10K in about 45:20
The Riegel formula, which you will find in the pace calculator, gives a 10K of about 45:19 and a 5K of about 21:44 for a 1:40 half. Without a result in that area, aim first for a sub-45 minute 10K. The shorter distance will make this pace familiar sooner. In training you can check it another way: 10–12 km (6–7.5 miles) of work at 4:44/km in one session, with short breaks, should end with something in reserve.
At 1:45 the pace is 4:59/km (8:01/mile), so you are 15 seconds per kilometer (23 per mile) short of 1:40. A 1:30 adds a fifth running day, and your threshold has to sit near 4:16/km (6:52/mile).
Frequently asked questions
Is 1:40 a good half marathon time?
Yes, a very good one. In the Vienna half marathon from 2006 to 2018 the fastest quarter of men averaged about 1:37 and the fastest quarter of women about 1:51. At 1:40 a man runs close to that fastest quarter, and a woman well inside it. It usually takes more than one training block.
What pace is a 1:40 half marathon?
4:44/km or 7:38/mile. At even pace you pass 5 km in 23:42, 10 km in 47:24 and 10 miles in 1:16:17.
Can I go from 1:50 to 1:40 in one training block?
Rarely. It is 10 minutes, or 28 seconds per kilometer (46 per mile), and in a single block of 12–16 weeks a jump like that mostly happens to people who are only starting to work on their threshold. More often it takes two or three seasons of smaller steps, with 1:50 as a stage on the way.
How many miles a week do I need for a sub-1:40 half marathon?
Take a peak of 55–75 km (34–47 miles) a week as the reference. At this level the quality of the Threshold sessions decides more than another 10 km of easy running.
Is a 45-minute 10K enough for a 1:40 half marathon?
Only just. From 45:00 the Riegel formula gives a half marathon of about 1:39:17, but it assumes you are prepared for the longer distance. On about 30 km (19 miles) a week the real result will be slower.
How long should my longest run be before a 1:40 half marathon?
The reference is 20–23 km (12.5–14.5 miles) about 2–3 weeks before the race. At an easy 5:55–6:15/km that takes roughly 2:00–2:25, longer than the race itself. Late in the block, finishing the long run close to 4:44/km does more than extra easy kilometers.
Is a 1:40 half marathon enough for a 3:30 marathon?
On paper, yes: the Riegel formula gives about 3:28:30. Most recreational runners slow down more than the formula assumes, and with the exponent typical of a large British sample the same half marathon gives about 3:37. Treat 3:30 as the upper limit. More on the sub-3:30 marathon page.


