1. PaceRunner
  2. Free tools

Race time predictor: what your 5K says about your 10K, half and marathon

A 25:00 5K predicts about 51:53 for a 10K, 1:55:04 for a half marathon and 3:57:55 for a marathon, if you've trained for the distance. Type any recent race below and the predictor gives the equivalent time and pace at every other distance, from the same equations PaceRunner uses for your training paces.

Your race

As m:ss or h:mm:ss, from a race or an all-out time trial in the last few months.

Type a time to see what it predicts for the other races.

Send workouts at these paces to your Garmin

How the prediction works

The predictor uses the Daniels–Gilbert equations, published by Jack Daniels and Jimmy Gilbert in the 1970s from measurements of trained runners. They describe two things: how much oxygen it costs to run at a given speed, and what share of your maximum you can sustain for a given time (close to all of it for 10 minutes, a little under 80% for three hours). A race result fixes one number, an estimate of your running fitness, and that number gives the time at which every other distance would cost the same effort.

It's the calculation behind PaceRunner's own paces: your profile takes a recent race, works out that fitness number, and sets easy, threshold, interval and race paces from it. The training pace calculator shows those paces for any time you type; this page shows the race times.

Predictions from common 5K times

From the Daniels–Gilbert equations, rounded to the second. Paces are on the predictor above.
5K10KHalf marathonMarathon
18:0037:201:22:332:52:25
20:0041:281:31:503:11:17
22:0045:371:41:073:30:03
25:0051:531:55:043:57:55
28:0058:112:08:574:25:28
30:001:02:232:18:104:43:39
35:001:12:532:40:595:28:30

Compared with Riegel's formula

The other prediction you'll meet is Pete Riegel's, from the late 1970s: the new time is the old time multiplied by the ratio of the distances raised to the power 1.06. It's one line of arithmetic and lands close to the equations for most runners, a little more cautious at the longer distances. Both start from the same 25:00 5K here:

RaceDaniels–GilbertRiegel
Mile7:247:31
10K51:5352:07
Half1:55:041:55:00
Marathon3:57:553:59:47

When two methods agree to within a minute or two, the disagreement isn't what decides your race. Training for the distance is.

Why your marathon may come out slower

The equations assume you're as well trained for the race you're predicting as for the one you ran. That holds well between a 5K and a 10K. It holds less well for a marathon, where many recreational runners finish slower than their 5K predicts, for reasons the 5K never tested:

  • Endurance. A marathon needs months of long runs and steady weekly volume. Without them, the last 10 km slow down whatever your 5K says.
  • Fueling. Past about 90 minutes, carbohydrate runs short unless you take it in on the way. A 5K never asks.
  • Heat, hills and wind. A warm day or a hilly course can cost several minutes over a marathon, and very little over a 5K.
  • Pacing. Starting a minute a mile too fast feels fine for 10 km and costs far more than a minute by the finish.

The closer the input race is to the one you're predicting, the better. A half marathon predicts a marathon better than a 5K does: 4:07:45 from a 2:00:00 half, which is 5:52 /km (9:27 /mi).

Getting a good input

  • Use a race or an all-out time trial from the last six to eight weeks. Fitness changes, and an old personal best overstates where you are now.
  • A flat, accurately measured course is best. A time from your watch on a hilly loop can be off by more than the method itself.
  • A 5K time trial works well: it's short enough to run hard on any weekend, and long enough to say something about endurance.

Questions

How accurate are race time predictors?
For neighboring distances, such as 5K to 10K, they're usually within a minute or two for a runner trained for both. For the marathon they're often optimistic, because the 5K doesn't test endurance or fueling.
What marathon time does a 20-minute 5K predict?
About 3:11:17 by the Daniels–Gilbert equations, or 4:32 /km (7:18 /mi), assuming marathon-specific training.
Is this the same as VDOT?
It uses the same published equations by Daniels and Gilbert. VDOT is a trademarked name for one product built on them, so we don't use it. The fitness number itself is on the VO2max calculator.
Which race should I put in?
The most recent all-out race or time trial, ideally the closest in distance to the race you're predicting. A half marathon predicts a marathon better than a 5K does.

Checked November 18, 2026