Predict target time for other distances from one running result – 5 km, 10 km, half marathon, and marathon using the Riegel formula.
Available
Inputs
Enter values
Live
Known Performance
Distance and time run.
km
Only fill in for custom distance.
h
min
s
Target
Distance for the prediction.
km
Only fill in for custom target distance.
Load example
For illustration only – your own calculation is what counts.
Explanation
How Running Prediction Calculator works
The running prediction calculator transfers a known running result to other distances. The basis is the Riegel formula, which represents the pace loss over longer distances with a fatigue exponent.
The Riegel Formula
Time predictionTarget time = Base time × (Target distance ÷ Base distance) ^ 1.06
The ratio of distances is raised to the power of the fatigue exponent 1.06 and multiplied by the base time.
Practice
Practical examples
For illustration only – your own calculation is what counts.
Half marathon from 10 km time
A 10 km run is completed in 50:00; the realistic half marathon time is sought.
Result: The Riegel formula predicts around 1:50:19 for the half marathon, slightly slower than double 10 km pace.
Marathon from half marathon
A half marathon is run in 1:45:00; the marathon prediction is sought.
Result: For the marathon, a prediction of around 3:38:55 results – the typical slowdown over double the distance.
Notes
Common mistakes
Expecting prediction without appropriate training
A 10 km personal best does not guarantee a marathon time if the long runs are missing.
Use a base time close to the target distance and train the target distance specifically.
Using outdated base time
Old or half-completed results provide false predictions.
Use a current, fully run race result.
FAQ
Frequently asked questions
What is the Riegel formula?
Peter Riegel described a relationship in 1981 for predicting time for another distance from a known running time: Target time = Base time × (Target distance ÷ Base distance) to the power of 1.06. The exponent 1.06 represents the fact that pace decreases with increasing distance.
Why the exponent 1.06?
An exponent of 1.0 would assume constant pace. The value 1.06 reflects real fatigue: the longer the distance, the slower the average pace. The value fits well for distances from a few minutes to several hours.
How accurate is the prediction?
It is most reliable when base and target distances are not too far apart and you are similarly trained for both. Extrapolating a 5 km time to the marathon is rougher than from 10 km to the half marathon.
Does the prediction replace specific training?
No. The formula assumes appropriate preparation. Those who have never trained over 10 km will generally not achieve the predicted marathon time. Base and distance training remain crucial.
Are course and weather considered?
No. Elevation, surface, heat, wind, and daily form are not included. The prediction is valid for comparable, flat, and fair conditions.
Which base time should I use?
Best is a current, fully run race result over a distance close to your target. A too-short or untrained test run distorts the prediction.
Limits
Limitations
Far outside the input distance, the prediction loses accuracy.
Elevation, surface, weather, and daily form are not considered.
Sources
Sources and references
Riegel, P. (1981): Athletic Records and Human EnduranceAmerican Scientist 69(3):285–290
Documents the prediction formula T2 = T1 × (D2 ÷ D1) ^ 1.06 with the fatigue exponent 1.06.
Retrieved: 09/23/2026 · Verified on: 09/23/2026
World Athletics – Competition DistancesWorld Athletics
Defines official race distances including half marathon and marathon.
View sourceRetrieved: 09/23/2026 · Verified on: 09/23/2026
Classifies the time by age and reference category.
Choose your course
You set the course.
Rechnerpilot counts views and actions anonymously – without cookies and without recognition – and improves through that regardless of your choice. In addition, you can allow Google Analytics, which is loaded only after your consent. You can change your choice at any time.
Your calculation inputs and results are transmitted on neither path.