Endurometry
A predictor scales a race you already ran; it does not invent fitness. Trust the Daniels time first, Riegel 1.06 as a close check, and 1.10 when the target is much longer. 10K in 48:30 → half 1:47:32 (Daniels), not a linear 1:42.
Last updated October 5, 2026
Multiplying a 10K pace by 4.2195 does not give a marathon. Fatigue grows faster than distance. The race time predictor implements two published ways to add that cost. This page is the model; turning a number into a marathon goal is estimate a marathon finish.
Three numbers, one race
Enter a result and a target. You get Daniels–Gilbert (the large figure), Riegel with exponent 1.06, and the same power law with 1.10. Same inputs as the calculator’s default: 10K in 48:30 (VDOT ≈ 41.5).
| Target | Daniels | Riegel 1.06 | 1.10 |
|---|---|---|---|
| 5K | 23:23 | — | 22:38 (faster) |
| Half (21.0975 km) | 1:47:32 (5:06/km) | 1:47:01 (5:04/km) | 1:50:15 (5:14/km) |
| Marathon | 3:42:54 (5:17/km) | 3:43:07 | 3:56:20 (5:36/km) |
A linear 48:30 × 2.10975 ≈ 1:42:17 half is too fast. The models add about five minutes because you cannot hold 10K pace for 21 km.
Riegel’s power law
T2 = T1 × (D2 / D1)^k
Peter Riegel (1981) fitted world-best performances. k = 1.06 is the usual road value. k = 1.10 stretches the change both ways: longer targets get slower, shorter ones faster. That is why 1.10 looks “better” at 5K and worse at the marathon. Use 1.10 as a conservative check for a longer race, not as a 5K goal. Equations: methodology.
Daniels–Gilbert and VDOT
Jack Daniels and Jimmy Gilbert (1979, Oxygen Power) start from oxygen cost and how much of VO2max you can hold for a given time. A known race fixes a VDOT (an endurance index, not a laboratory VO2max). The predictor then searches for a target time whose cost matches what that VDOT can sustain. Daniels usually sits between the two power laws — that is why it is the headline.
The same VDOT drives the training pace calculator. Two runners with the same 10K share a VDOT even if their lab tests differ.
When the estimate is weak
5K → 10K is more trustworthy than 5K → marathon. The models assume a genuine race effort, similar conditions, and that you have trained for the target. A downhill 5K or an easy 10K will lie. Heat, hills, altitude and a first-time distance are not in the formula. Turn a trusted time into even splits with the pace calculator.
Related Tools
Race Time Predictor
Predict your finish time at any distance using the Riegel power law and the Daniels–Gilbert VO2max model.
Training Pace Calculator (VDOT Zones)
Get your VDOT from a recent race and see easy, marathon, threshold, interval and repetition pace zones plus heart-rate zones.
Running Pace Calculator
Enter distance and finish time to get your pace per km and mile, average speed and per-kilometre splits.
Related guides
Estimate a marathon finish time
A practical workflow: scale a recent 10K or half with Daniels, add time if you have not trained the distance, then turn the finish into even-split pace.
5K, 10K and half marathon pace
How to set 5K, 10K and half goals from another race, why even splits still win, and why you must not start a 5K like a 400 m.
How the numbers are calculated
Race time prediction · Rounding and displayed precision · Sources
Frequently Asked Questions
Which number should I use?
Daniels first. Riegel 1.06 should be close from 5K to marathon. Read 1.10 when the target is much longer than the race you already ran.
Is VDOT the same as VO2max?
No. VDOT is fitted from a race. It is useful for scaling and training paces; it is not a lab measurement.
Can I predict from a training run?
You can type it, but the models assume race effort. Easy runs make every target too slow; downhill time trials make them too fast.
Why is 1.10 sometimes faster?
A steeper exponent amplifies the distance change in both directions. From a 10K, 1.10 5K is quicker; 1.10 marathon is slower.
Want a new tool or fix?
Tell us which calculator, pace chart or fix would help — we build what runners ask for. Replies usually come back in 1–2 days.