Free tool
Plug in a recent finish time and get predictions for every other distance.
I ran a
in
mm:ss or h:mm:ss
Weekly mileage (optional)
This gives you a snapshot. athletedata connects to Garmin, Apple Health, WHOOP, and Oura to give you an AI coach that builds race plans around your actual training and recovery data.
Try free for 7 days →The predictions use the Riegel formula (T2 = T1 × (D2/D1)1.06), the standard race equivalence model since 1977. The 1.06 exponent captures how pace degrades as distance increases.
We add a mileage adjustment for half marathon and marathon predictions. A 22-minute 5K runner doing 50 miles/week will run a very different marathon than one doing 15. The formula alone can't account for that, so we nudge the prediction based on your training volume.
For best results: use a recent all-out race effort (not a training run), and predict to distances close to your input. A 5K → 10K prediction is very reliable. A 5K → marathon prediction on 15 miles/week will be optimistic.
Within 2-3% for well-trained runners with adequate mileage. If you're under 30 miles/week, expect your actual time to be a bit slower than predicted.
You can, but treadmill times run fast (no wind, flat surface, slight belt assist). A recent outdoor race gives the most accurate predictions.
Most plans peak at 30-50 mi/week. Beginners around 30-35, intermediate 40-50, advanced 55+. Higher mileage closes the gap between your short-race fitness and marathon performance.
The marathon is disproportionately harder. A 1:45 half predicts roughly a 3:40 marathon, not 3:30. That's the physiology of running 26.2 miles - the Riegel exponent captures this.