Marathon Predictor — 5K, 10K & Half Marathon Race Predictor

A marathon predictor is a calculator that estimates your marathon finish time from a recent shorter race — typically a 5K, 10K, or half marathon result. The same engine doubles as a 5K race predictor, half marathon predictor, or general race time predictor — you pick the known distance, you pick the target distance, and the math runs in either direction.

The most reliable predictors run multiple scientific models so you see a confidence range rather than a single number. This calculator runs three at once:

• Riegel formula (1981) — the most widely used race time predictor, applying a fatigue exponent of 1.06 to scale performance across distances. Most optimistic of the three.
• Cameron formula (1999) — distance-specific adjustments that account for fatigue accumulating non-linearly across distances. Typically more conservative for longer predictions.
• Daniels VDOT — converts your race performance to a VO2max-equivalent score, then maps that score to the target distance using exercise physiology. Generally considered the most accurate marathon time predictor by exercise physiologists.

For accuracy, use a known race from within the past 8-12 weeks at a distance close to your target. A 10K predicting a half marathon is highly reliable; a 5K predicting a marathon is more of a ballpark estimate that you should pair with marathon-specific endurance training.

The Riegel formula, published by Peter Riegel in 1981, is remarkably accurate for predicting race times between similar distances — for example, from a 10K to a half marathon. Studies have shown it predicts within 1-3% for well-trained runners when the distance ratio is less than 4:1. However, its accuracy decreases for large jumps such as predicting a marathon from a 5K, where it tends to be overly optimistic because it uses a fixed fatigue exponent (1.06) that does not fully capture the exponential energy demands of longer distances.

For best results, use a known race result from within the past 3 months and choose a known distance as close to your target distance as possible.

Yes, but with important caveats. Predicting a marathon from a 5K represents an 8.4x distance increase, which introduces significant uncertainty. The Riegel formula will produce the most optimistic estimate, while Cameron and Daniels/VDOT models tend to give more conservative — and often more realistic — predictions for such large jumps.

The main issue is that marathon performance depends heavily on factors that 5K races do not test: glycogen depletion, fat oxidation efficiency, mental endurance beyond 2 hours, and fueling strategy. A runner with a 20-minute 5K might predict a 3:12 marathon using Riegel, but realistically finish closer to 3:20-3:35 depending on their long-run training volume.

For the most reliable marathon prediction, use a half marathon or at minimum a 10K result.

Each model approaches race prediction from a different angle:

• Riegel (1981) uses a simple power law: T2 = T1 x (D2/D1)^1.06. The exponent 1.06 represents the universal fatigue factor. It is elegant and widely used but assumes a constant fatigue rate across all distances.
• Cameron (1999) uses distance-specific adjustment factors derived from polynomial equations. It accounts for the fact that fatigue accumulates differently at different distances — the slowdown from 5K to 10K is proportionally less than from half marathon to marathon. This makes it more conservative for longer predictions.
• Daniels/Gilbert VDOT converts your race performance to a VO2max-equivalent score (VDOT), then uses exercise physiology equations to predict what time that aerobic capacity could sustain at another distance. It accounts for both the oxygen cost of running and the decreasing fraction of VO2max you can maintain as duration increases.

When all three models agree closely, you can be confident in the prediction. When they diverge, the Daniels model is generally considered most reliable by exercise physiologists.

To run a sub-4 hour marathon, you generally need a 5K time of approximately 25:00 or faster. Here is how the three models translate specific 5K times to marathon predictions:

• 23:00 5K: Riegel predicts ~3:41, Daniels ~3:45 — comfortably sub-4
• 25:00 5K: Riegel predicts ~4:00, Daniels ~4:05 — borderline sub-4
• 27:00 5K: Riegel predicts ~4:19, Cameron ~4:25 — likely above 4:00

However, these are best-case predictions assuming adequate marathon-specific training. A fast 5K demonstrates aerobic capacity, but sub-4 also requires sufficient long run volume (peaking at 30-35 km), proper fueling strategy, and pacing discipline. Many runners with a 26-minute 5K miss sub-4 in their first marathon due to insufficient endurance base.

You can, but expect the predictions to be less accurate. Race conditions — adrenaline, competition, tapered rest, ideal pacing — typically produce times 3-5% faster than solo training efforts. If you input a training time, the predictor will underestimate your potential.

To get usable results from training data:

• Use a time trial run at maximum effort on a measured, flat course
• Subtract approximately 1-2% from your training time to approximate race effort
• Alternatively, use a training pace calculator to work backwards from your workout paces to an equivalent race fitness level

Parkrun results are an excellent free alternative — they provide genuine 5K race conditions every week.

The simplest method is to double your half marathon time and add 10-20 minutes — this rough rule accounts for the exponential fatigue factor in longer distances. For example, a 1:45 half marathon suggests a marathon time of approximately 3:40-3:50.

For a more precise prediction, our calculator uses three scientific models:

• Riegel formula: multiplies by a distance-ratio factor with exponent 1.06, typically giving the most optimistic prediction
• Cameron formula: uses distance-specific coefficients, often predicting slightly slower than Riegel
• Daniels/VDOT: derives your VO2max equivalent from the half marathon and maps it to marathon performance

Half marathon to marathon is the most reliable prediction path because the physiological demands overlap significantly. Research shows only 5% of runners beat their Riegel-predicted marathon time from a half marathon — so treat the average of all three models as a realistic ceiling. Not sure if your half time itself is strong for your age? See average half marathon times by age.

Race prediction formulas were developed from large datasets of experienced, well-trained runners. Several factors can cause real-world marathon times to be slower than predicted:

• Insufficient long-run training: predictions assume you have the endurance base to sustain the effort. Without regular 25-35 km training runs, you will likely hit the wall.
• First marathon: first-time marathoners tend to run 10-15% slower than predicted from shorter race times due to inexperience with pacing, fueling, and the psychological demands beyond 30 km.
• Course difficulty: predictions assume a flat, fast course. Hills, wind, or trail surfaces add time. Use our elevation adjustment feature for hilly courses.
• Weather: temperatures above 15C progressively slow marathon performance. Our temperature adjustment accounts for this — enter the expected race-day temperature for a more realistic estimate.

If your predicted time seems ambitious, target the Daniels/VDOT result or add 5-10% as a buffer for your first attempt at the distance.

Heat significantly impacts distance running performance, especially in races lasting over 90 minutes. Our calculator includes a temperature adjustment feature — enter the expected race-day temperature to see adjusted predictions.

General guidelines for heat impact on marathon performance:

• 10-12C (50-54F): optimal range, no adjustment needed
• 15C (59F): approximately 2% slowdown
• 20C (68F): approximately 5% slowdown
• 25C (77F): approximately 10% slowdown
• 30C (86F): approximately 16% slowdown

These figures are based on research by Ely et al. (2007) analyzing weather impacts on marathon performance. The effect is more pronounced for slower runners who are exposed to heat for longer durations. If your target race is in warm conditions, reduce your goal pace accordingly rather than attempting your cool-weather predicted time.

VDOT is a concept developed by legendary running coach Jack Daniels and mathematician Jimmy Gilbert in their 1979 research. It represents your current running fitness as a single number equivalent to your VO2max, but derived entirely from race performance rather than laboratory testing.

The VDOT system works by combining two physiological relationships: the oxygen cost of running at a given pace (which increases with speed) and the fraction of VO2max that a runner can sustain over a given duration (which decreases as race time increases). By knowing your race time at one distance, the model calculates your VDOT, then uses it to predict performance at any other distance.

A typical recreational runner might have a VDOT of 35-45, while elite marathoners score 70-85. The beauty of the system is that a VDOT of 50 predicts very specific times for every distance, making it a powerful training and prediction tool.

For well-trained runners predicting one distance step up (e.g., 10K to half marathon), predictions are typically accurate within 2-3% of actual race time. Predicting across two or more distance steps (e.g., 5K to marathon) reduces accuracy to roughly 5-8%.

Factors that reduce accuracy include: insufficient long-run training for the target distance, extreme weather conditions, hilly courses, poor race-day nutrition, and significant changes in fitness since the input race. First-time marathoners tend to run 10-15% slower than predicted from 5K or 10K times, because marathon-specific endurance requires dedicated training beyond aerobic fitness.

To improve prediction accuracy, use a recent race (within 8-12 weeks) at the closest available distance to your target. A 10K result will predict a half marathon more accurately than a 5K result would.