The formula: T2 = T1 × (D2/D1)^1.06
T1 and D1 are the known time and distance from a recent race; D2 is the new target distance. The ratio of the two distances is raised to the power 1.06, rather than simply 1, because pace realistically slows as distance increases — a straight-line scaling would overestimate how fast someone could sustain a much longer effort.
A worked example: predicting a 10K from a 5K
A 5K completed in 25:00 (25 minutes) predicts a 10K time of 25 × (10/5)^1.06 ≈ 52.12 minutes, formatted as 0:52:07. Note this isn't simply double the 5K time (which would be 50:00) — the 1.06 exponent adds a small additional slowdown for the longer distance.
Chaining the prediction further, from 10K to marathon
Feeding that predicted 10K result (52 minutes, 7 seconds) back in as the known time, with a target of 42.195 km, predicts a marathon time of 52.1167 × (42.195/10)^1.06 ≈ 239.75 minutes, formatted as 3:59:45.
Why the marathon prediction is less trustworthy
This marathon estimate is now built on a prediction of a prediction, spanning a much larger distance gap (5K to marathon is more than 8 times the original distance) than the formula was validated for. Riegel's formula holds up best between similar distances — the 5K-to-10K step is reasonably reliable, but chaining it all the way to a marathon compounds both the formula's own approximation and the uncertainty of predicting from a predicted number rather than an actual race result.