How to Predict Your Race Time

One race predicts another

A 10K in 45:00 predicts a half marathon of about 1:39:17 — a target pace of 4:42/km.

That comes from Riegel's formula, published by Peter Riegel in Runner's World in 1977:

T₂ = T₁ × (D₂ ÷ D₁)^1.06

Everything is in the 1.06

If the exponent were exactly 1, your pace would be identical at every distance. No runner experiences that.

At 1.06, doubling the distance multiplies your time by 2^1.06 = 2.085 rather than 2. So your pace slows by about 4.25% every time the distance doubles.

That one number is the entire content of the formula. Riegel derived it empirically from race data — it's a statement about how race performance scales.

Don't confuse that with the tool's default. 10K → half marathon isn't a doubling — it's a 2.11× jump — so the calculator shows 2.206× and 4.58%, not 2.085 and 4.25%. The doubling figures are the useful ones to carry in your head; the tool gives you the exact ratio for whatever two distances you actually enter.

Units cancel. Because the calculation depends only on the ratio D₂/D₁, kilometres and miles give identical answers — provided both boxes use the same one. Mixing them is the only way to get it wrong, and nothing will warn you.

The assumption that breaks

Here's the part worth being blunt about.

The formula assumes you're equally well trained for both distances.

A 5K time predicts a marathon on the basis that you've done marathon training — the long runs, the fuelling practice, the hours on your feet. If you haven't, the prediction is describing a runner who doesn't exist yet.

Riegel tells you how race performance scales with distance. It knows nothing about whether you've put in the work. That's why the number so often looks flattering: it's the time a marathon-trained version of you could run.

Accuracy falls off with the size of the jump

Prediction	Reliability
5K → 10K	Good
10K → half	Reasonable
Half → marathon	Rough guide
5K → marathon	A guess with a decimal point

Published assessments put these formulas at roughly 80% accuracy — about one runner in five misses meaningfully.

The exponent is also an average. Stronger endurance runners slow less than 1.06 predicts; newer runners typically slow more.

Which means your own results beat it. If you've honestly raced two distances, the exponent that fits your two times describes you better than Riegel's does — and knowing whether you fade more or less than average is worth more than any single prediction.

Where it stops working

The formula holds for race durations between roughly 3.5 and 230 minutes.

Past the marathon it becomes unreliable for reasons that aren't physiological scaling at all: terrain, elevation, nutrition, sleep deprivation, and simply being awake for a very long time. No exponent describes those.

Predicting downwards

It runs backwards perfectly well — the pace-change figure just comes out negative, meaning you'd expect to run faster per kilometre over the shorter distance.

It tends to be less useful that way, though. A marathon time reflects endurance and fuelling more than the raw speed a 5K demands, so predicting down usually flatters you less than predicting up flatters you.

Reading the outputs

The predicted finish is shown to the nearest minute, because that is honestly the precision the formula has — but the exact figure in seconds sits underneath it if you want to work backwards yourself.

The target pace comes as two numbers, minutes and — and seconds, read together: 4 and 42 means 4:42 per kilometre. Splitting it that way is deliberate — nobody paces a race off "4.7 minutes per km", and working the seconds out from a single rounded total is how a pace ends up displayed as 4:60.

Enter your benchmark time the same way, in hours, minutes and seconds. On a 5K especially, dropping the seconds moves the prediction more than you would expect: ten seconds on a 20-minute 5K shifts a predicted marathon by over a minute and a half — because the formula multiplies your time by more than eight to get there, so it multiplies your rounding too.

Use it to set a pace, not a promise

The genuinely useful output isn't the finish time — it's the target pace. 4:42/km is something you can hold in your head and check at each marker; 1:39:17 is something you find out about at the end.

Take the pace, plan to go through halfway on it rather than ahead of it, and have a slower fallback for heat, wind or a bad day.

Print the race prediction sheet. The column worth filling in is the one where you write what actually happened — after two or three races it tells you more about yourself than the formula does.

Not medical advice. Training decisions, injury and whether to attempt a distance at all are matters for you and a qualified coach or clinician.

Sources: RunnersConnect — how accurate are race calculators · Peter Riegel