Race time predictor
One recent race in, four predicted finish times out — computed with Riegel's formula, the same one behind almost every predictor on the internet, with the exponent shown so you can check the working.
Enter a recent race distance and time.
Read the marathon number with suspicion. Predicting 42 km from a race of 10 km or less assumes an endurance base the short race cannot demonstrate. Treat it as a ceiling, not a target.
Predicted times from a 5K
Server-rendered from the same formula the calculator uses. Read down to your 5K time and across.
| 5K time | Pace /km | 10K | Half marathon | Marathon |
|---|---|---|---|---|
| 15:00 | 3:00 | 31:16 | 1:09:00 | 2:23:52 |
| 16:00 | 3:12 | 33:22 | 1:13:36 | 2:33:27 |
| 17:00 | 3:24 | 35:27 | 1:18:12 | 2:43:03 |
| 18:00 | 3:36 | 37:32 | 1:22:48 | 2:52:38 |
| 19:00 | 3:48 | 39:37 | 1:27:24 | 3:02:14 |
| 20:00 | 4:00 | 41:42 | 1:32:00 | 3:11:49 |
| 21:00 | 4:12 | 43:47 | 1:36:36 | 3:21:25 |
| 22:00 | 4:24 | 45:52 | 1:41:12 | 3:31:00 |
| 23:00 | 4:36 | 47:57 | 1:45:48 | 3:40:36 |
| 24:00 | 4:48 | 50:02 | 1:50:24 | 3:50:11 |
| 25:00 | 5:00 | 52:07 | 1:55:00 | 3:59:47 |
| 26:00 | 5:12 | 54:12 | 1:59:36 | 4:09:22 |
| 27:00 | 5:24 | 56:18 | 2:04:12 | 4:18:58 |
| 28:00 | 5:36 | 58:23 | 2:08:48 | 4:28:33 |
| 29:00 | 5:48 | 1:00:28 | 2:13:24 | 4:38:09 |
| 30:00 | 6:00 | 1:02:33 | 2:18:00 | 4:47:44 |
| 31:00 | 6:12 | 1:04:38 | 2:22:36 | 4:57:19 |
| 32:00 | 6:24 | 1:06:43 | 2:27:12 | 5:06:55 |
| 33:00 | 6:36 | 1:08:48 | 2:31:48 | 5:16:30 |
| 34:00 | 6:48 | 1:10:53 | 2:36:24 | 5:26:06 |
| 35:00 | 7:00 | 1:12:58 | 2:41:00 | 5:35:41 |
| 36:00 | 7:12 | 1:15:03 | 2:45:36 | 5:45:17 |
| 37:00 | 7:24 | 1:17:09 | 2:50:12 | 5:54:52 |
| 38:00 | 7:36 | 1:19:14 | 2:54:48 | 6:04:28 |
| 39:00 | 7:48 | 1:21:19 | 2:59:24 | 6:14:03 |
| 40:00 | 8:00 | 1:23:24 | 3:04:00 | 6:23:39 |
Predicted times from a 10K
A 10K is a considerably better base for a half-marathon prediction than a 5K is, and a half marathon is better still for a marathon.
| 10K time | Pace /km | 5K | Half marathon | Marathon |
|---|---|---|---|---|
| 30:00 | 3:00 | 14:23 | 1:06:12 | 2:18:00 |
| 32:00 | 3:12 | 15:21 | 1:10:36 | 2:27:12 |
| 34:00 | 3:24 | 16:18 | 1:15:01 | 2:36:24 |
| 36:00 | 3:36 | 17:16 | 1:19:26 | 2:45:36 |
| 38:00 | 3:48 | 18:14 | 1:23:51 | 2:54:48 |
| 40:00 | 4:00 | 19:11 | 1:28:15 | 3:04:00 |
| 42:00 | 4:12 | 20:09 | 1:32:40 | 3:13:13 |
| 44:00 | 4:24 | 21:06 | 1:37:05 | 3:22:25 |
| 46:00 | 4:36 | 22:04 | 1:41:30 | 3:31:37 |
| 48:00 | 4:48 | 23:01 | 1:45:54 | 3:40:49 |
| 50:00 | 5:00 | 23:59 | 1:50:19 | 3:50:01 |
| 52:00 | 5:12 | 24:56 | 1:54:44 | 3:59:13 |
| 54:00 | 5:24 | 25:54 | 1:59:09 | 4:08:25 |
| 56:00 | 5:36 | 26:52 | 2:03:34 | 4:17:37 |
| 58:00 | 5:48 | 27:49 | 2:07:58 | 4:26:49 |
| 1:00:00 | 6:00 | 28:47 | 2:12:23 | 4:36:01 |
| 1:02:00 | 6:12 | 29:44 | 2:16:48 | 4:45:13 |
| 1:04:00 | 6:24 | 30:42 | 2:21:13 | 4:54:25 |
| 1:06:00 | 6:36 | 31:39 | 2:25:37 | 5:03:37 |
| 1:08:00 | 6:48 | 32:37 | 2:30:02 | 5:12:49 |
| 1:10:00 | 7:00 | 33:34 | 2:34:27 | 5:22:01 |
| 1:12:00 | 7:12 | 34:32 | 2:38:52 | 5:31:13 |
| 1:14:00 | 7:24 | 35:30 | 2:43:16 | 5:40:25 |
| 1:16:00 | 7:36 | 36:27 | 2:47:41 | 5:49:37 |
| 1:18:00 | 7:48 | 37:25 | 2:52:06 | 5:58:49 |
| 1:20:00 | 8:00 | 38:22 | 2:56:31 | 6:08:01 |
The formula, and its honest limits
Peter Riegel, an engineer and a course measurer, published this in American Scientist in 1981:
T₂ = T₁ × (D₂ ÷ D₁)1.06 — where T₁ is a known time over distance D₁, and T₂ is the predicted time over D₂.
The exponent of 1.06 is what makes it work. If human endurance were perfect the exponent would be 1.0 and doubling the distance would exactly double the time. It is slightly above 1, which encodes the fact that pace decays as distance grows — and it decays by a remarkably consistent amount across a very large population of runners.
Three things it cannot know, and all three matter more in India than in the countries where the formula was fitted:
- Whether you trained for the distance. The formula assumes appropriate preparation. A 22-minute 5K predicts a 3:33 marathon; a runner whose longest run is 15 km will not run that, and no formula can tell the difference between the two.
- The weather. A prediction is a cool-morning number. A humid January start in Mumbai or an August marathon in Hyderabad costs minutes that appear nowhere in the equation.
- The course. Flat Ahmedabad and hilly Hyderabad get the same prediction and will not produce the same result.
Use it the way it deserves to be used: as a sanity check on a goal, not as the goal. Then take the pace it implies into the split calculator and see whether the per-kilometre number looks like something you have actually held in training.
Questions
How accurate is a race time predictor?
Riegel's formula is accurate to within a few per cent when you are predicting a distance close to the one you raced and you have actually trained for the longer distance. Predicting a marathon from a 5K is where it breaks down: the formula assumes endurance you may not have built, and it will hand you a marathon time you cannot run.
What is the Riegel formula?
T2 = T1 × (D2 ÷ D1) raised to the power 1.06, where T1 is your known time over distance D1. Peter Riegel published it in American Scientist in 1981 after fitting the curve to a large set of race results. The exponent 1.06 says that every time you double the distance, your time slightly more than doubles.
Can I predict a marathon from a half marathon?
More safely than from any shorter distance, but still with a caveat. Riegel gives roughly your half time multiplied by 2.09. That is a fair target if your training included long runs of 30 km or more; if your longest run was 18 km, the formula is describing a runner you are not yet.
Why is my predicted time faster than my actual race?
Usually one of three things: the prediction came from too short a race, the training did not support the distance, or the day was hot. Heat alone routinely costs several minutes over a marathon, and a prediction formula knows nothing about weather.
Every time, pace and physiological figure on this site is computed from a published formula, and the formula is named and shown on the method page. Nothing here is predicted by a model you cannot check. Race details change — always confirm dates, routes and entry with the official organiser. Training guidance is general information, not individual coaching or medical advice.