How the race time predictor works
Every runner slows down as races get longer, but they slow down in a remarkably predictable way. If you know how fast you can run one distance, you can make a solid estimate of what you are capable of over another. This calculator runs your result through two of the most widely used prediction models and shows both, so you can see a realistic range rather than a single number.
- Pick the distance you raced and enter your finish time. Use chip time from a certified course if you have it.
- Choose an endurance profile. Leave it on “Standard” unless you know you are much stronger at short races (choose low mileage) or long races (choose high mileage).
- Compare the Riegel and VDOT columns. Where they agree, the prediction is robust. Where they differ, the truth usually lies between them.
The Riegel formula
In 1977 research engineer Peter Riegel noticed that the relationship between race time and distance in running — and in swimming, cycling and walking — follows a power law. His formula, first published in Runner’s World and later in American Scientist (1981), is still the most widely used race predictor in the world:
T₁ and D₁ are your known time and distance, D₂ is the distance you want to predict, and 1.06 is the fatigue exponent. If there were no fatigue at all, the exponent would be 1.00 and you would run every distance at the same pace. The extra 0.06 captures how much you slow down as the distance doubles — about 4 % in pace.
Worked example
A 25:00 5K predicting a 10K: 25:00 × (10 ÷ 5)1.06 = 1,500 s × 2.085 = 3,127 s = 52:07. For the half marathon: 1,500 × (21.0975 ÷ 5)1.06 = 1,500 × 4.600 = 6,900 s = 1:55:00.
The VDOT model
The second column uses the equations published by Jack Daniels and Jimmy Gilbert in Oxygen Power (1979). Instead of a single exponent, VDOT combines two physiological curves: the oxygen cost of running at a given speed, and the percentage of your maximal aerobic capacity you can sustain for a given duration. Your race result is converted into a single fitness score (your VDOT) and then projected onto every other distance.
In practice the two models agree closely from about 3K to the half marathon. VDOT tends to be slightly more generous for the marathon and slightly harsher for very short races.
Race equivalency chart: 5K to marathon
This table shows VDOT-equivalent performances for common 5K times. Read across to see what a given 5K suggests you could run over longer distances if you are equally well trained for them.
| 5K | VDOT | 10K | Half marathon | Marathon |
|---|---|---|---|---|
| 16:00 | 64.6 | 33:13 | 1:13:19 | 2:33:26 |
| 17:00 | 60.2 | 35:16 | 1:17:56 | 2:42:56 |
| 18:00 | 56.3 | 37:20 | 1:22:33 | 2:52:25 |
| 19:00 | 52.9 | 39:24 | 1:27:11 | 3:01:52 |
| 20:00 | 49.8 | 41:28 | 1:31:50 | 3:11:17 |
| 21:00 | 47.0 | 43:32 | 1:36:28 | 3:20:41 |
| 22:00 | 44.5 | 45:37 | 1:41:07 | 3:30:03 |
| 23:00 | 42.3 | 47:42 | 1:45:46 | 3:39:22 |
| 24:00 | 40.2 | 49:48 | 1:50:25 | 3:48:40 |
| 25:00 | 38.3 | 51:53 | 1:55:04 | 3:57:55 |
| 26:00 | 36.6 | 53:59 | 1:59:42 | 4:07:08 |
| 27:00 | 35.0 | 56:05 | 2:04:20 | 4:16:19 |
| 28:00 | 33.5 | 58:11 | 2:08:57 | 4:25:28 |
| 30:00 | 30.8 | 1:02:23 | 2:18:10 | 4:43:39 |
| 32:00 | 28.5 | 1:06:35 | 2:27:20 | 5:01:42 |
| 35:00 | 25.6 | 1:12:53 | 2:40:59 | 5:28:30 |
| 38:00 | 23.1 | 1:19:11 | 2:54:30 | 5:55:01 |
| 40:00 | 21.7 | 1:23:22 | 3:03:27 | 6:12:33 |
Why marathon predictions are often too optimistic
Both models assume you are equally well prepared for every distance. For races up to the half marathon that assumption is reasonable for most runners. The marathon is different: after roughly 90–120 minutes of hard running, glycogen stores run low and muscles fatigue in ways that shorter races never reveal. Several factors decide whether you match your prediction:
Weekly mileage
Runners on low weekly volume usually slow down much more in the second half of a marathon than prediction models expect.
Long runs
Regular long runs of 2–3 hours build the muscular endurance and fat-burning capacity the marathon demands.
Fuelling
Taking in carbohydrate during the race delays glycogen depletion. Practise it in training so your stomach tolerates it.
Conditions & course
Heat, wind, hills and crowding can each cost minutes. Predictions assume a flat course in cool weather.
A practical rule of thumb. If you run less than about 50 km (30 miles) a week, add 5–10 % to your predicted marathon time and use that as your starting goal. Choose the Low mileage profile above for a similar adjustment.
Which race should you predict from?
- Use the longest recent race you have. A half marathon is a much better predictor of marathon time than a 5K, because it tests more of the same endurance.
- Use an all-out effort. Training runs, parkruns jogged with friends or races run in the heat will under-predict your ability.
- Keep it recent. Fitness changes within weeks. A result from last season describes the runner you were, not the runner you are.
- Masters runner? Compare results across ages with the age grading calculator.
- Check the course. Certified, flat courses give the most reliable input. Trail and hilly races will make predictions too slow.
Turning a prediction into a goal
A prediction tells you what your current fitness supports. To turn it into a race plan, take the time you believe in, then use the marathon pace calculator for splits, the VDOT calculator for the training paces that will get you there, and the pace band generator for race day. Browse our goal pace charts for ready-made plans for popular targets such as a sub-4 marathon or a sub-2 half.
- Riegel, P. S. (1981). Athletic records and human endurance. American Scientist, 69(3), 285–290.
- Daniels, J. & Gilbert, J. (1979). Oxygen Power: Performance Tables for Distance Runners.
Last reviewed: September 2026 · Health disclaimer