๐ฎ Race Time Predictor
One recent race result in, predicted times for every distance out.
Predicted Marathon Time
For informational purposes only, not medical advice. Consult a healthcare professional.
How to Use This Race Predictor
Peter Riegel, a research engineer and marathoner, found that race times scale predictably with distance: double the distance, add about 8.4 percent to the pace. Enter one recent all-out race result and get predicted times for 5K, 10K, half marathon, and marathon.
Enter your most recent race distance in km and your finish time. Use an all-out effort from the last 4 to 6 weeks.
Read the predictions. The marathon number is the headline; the shorter distances are pacing references.
Treat predictions as "with proper training" numbers. A 5K predicts the marathon you could run after a full marathon buildup, not next weekend.
Recalculate after every race. The trend of predictions is more informative than any single one.
The Riegel Formula
Riegel analyzed world records across distances and found the 1.06 exponent fit remarkably well: predicted time equals known time times the distance ratio raised to 1.06. A 25:00 5K (1500 s) predicts a 10K of 1500 x 2^1.06 = 1500 x 2.08496 = 3127.4 s = 52:07. The half marathon: 1500 x (21.0975/5)^1.06 = 6900.3 s = 1:55:00. The marathon: 1500 x (42.195/5)^1.06 = 14386.7 s = 3:59:47. The exponent is the whole insight: because it exceeds 1.0, pace fades as distance grows, about 8.4 percent slower per distance doubling.
Worked Example: 25:00 5K
Enter 5 km and 25:00. The calculator predicts 10K in 52:07, half marathon in 1:55:00, and marathon in 3:59:47. Verify the 10K by hand: 1500 x (10/5)^1.06 = 1500 x 2.08496 = 3127.4 s = 52:07. The predictions assume equivalent training for each distance. Most runners find Riegel slightly optimistic for the marathon, since few 5K specialists have the endurance base a marathon needs, and slightly pessimistic for well-trained marathoners moving down.
When Predictions Miss
Riegel assumes similar aerobic development across distances, which fails at the extremes. Sprinters predicting marathons get fantasy numbers; ultra-runners predicting 5Ks get sandbagged ones. Heat, hills, altitude, and wind all break the formula, which assumes equivalent fast conditions. Age-graded tables and VDOT (Jack Daniels) offer alternatives: VDOT tends to be kinder to endurance types, Riegel to speed types. Use predictions to set training paces and honest goals, not as promises. The best predictor of your next race remains your last few months of training.
Using Predictions in Training
Predicted times convert directly into workout paces: tempo pace near the predicted 10K-to-half-marathon effort, interval pace near predicted 5K effort, easy pace well slower. Many training plans (including Daniels) are built entirely on one recent race input. Recompute monthly during a buildup; if the predicted marathon drops from 4:05 to 3:55 over eight weeks, the training is working. If it stalls, adjust the plan, not the goal. And always sanity-check against how workouts feel: formulas guide, legs decide.
Second worked example: 52:00 10K
Enter 10 km and 52:00. Total = 3120 seconds. Half marathon = 3120 x (21.0975/10)^1.06 = 3120 x 2.2062 = 6883.3 s = 1:54:43. Marathon = 3120 x (42.195/10)^1.06 = 3120 x 4.6003 = 14,352.9 s = 3:59:13. Verify the half by hand: 21.0975/10 = 2.10975, raised to 1.06 = 2.2062, times 3120 = 6883.3 s = 114.72 min = 1:54:43. Compare with the 25:00 5K example, which predicted 3:59:47 for the marathon: nearly identical marathon predictions from different input distances, which is exactly what a good model should do. When two different races predict the same marathon, trust the number.
When they disagree, trust the longer race: a 10K predicts the marathon better than a 5K does, because it samples more of the endurance systems the marathon demands. A 5K-based 3:59 and a 10K-based 4:10 means your speed outruns your endurance, so train the long run, not the track.
Riegel vs VDOT vs McMillan
Three prediction systems dominate running. Riegel is pure math: one exponent fits all distances, no physiology, and it works surprisingly well for runners with balanced training. VDOT, from Jack Daniels, converts a race time into an aerobic fitness score, then reads equivalent performances off tables; it also generates full training paces, which is why coaches love it. The McMillan calculator uses regression on large race datasets and tends to give slightly more conservative marathon predictions than Riegel. In practice, Riegel runs a touch optimistic for the marathon, VDOT sits in the middle, and McMillan runs a touch conservative. Use all three: if they cluster within a few minutes, the goal is solid; if they spread 15 minutes apart, your training is lopsided and the conservative number is the honest one.
Common prediction mistakes
Mistake one is the stale personal best. A 5K PR from two years and 15 pounds ago predicts a marathon you cannot currently run; use races from the last 4 to 6 weeks. Mistake two is the downhill qualifier: a net-downhill half marathon predicts an uphill marathon badly, because the formula assumes equivalent courses. Mistake three is predicting across extremes. Riegel assumes similar aerobic development at both distances, so a miler's marathon prediction is fantasy and an ultrarunner's 5K prediction is sandbagging. Predict within one or two steps of distance, and treat the marathon number as a with-proper-training figure, not a next-weekend promise.
For informational purposes only, not medical advice. Consult a healthcare professional.
Worked Example: 25:00 5K to Marathon
A 25:00 5K predicts a marathon via the Riegel formula: T2 = T1 x (D2/D1)^1.06. Marathon/5K distance ratio = 42.195/5 = 8.439. 8.439^1.06 = e^(1.06 x ln 8.439) = e^(1.06 x 2.133) = e^2.261 = 9.59. Predicted marathon = 25:00 x 9.59 = 239.8 minutes = 3:59:48. So a 25-minute 5K runner with proper marathon training runs just under 4 hours. The 10K prediction: ratio 2, 2^1.06 = 2.085, 25:00 x 2.085 = 52:08.
The formula assumes equivalent training for the target distance, which is the catch. A 5K specialist jumping to the marathon untrained runs 4:30, not 3:59, because endurance, fueling, and pacing are untested. The prediction is a ceiling set by your demonstrated fitness, not a promise. Train for the distance and the number is realistic; skip the long runs and it is fiction.
When Predictions Miss
Riegel's exponent 1.06 is an average; individuals range from about 1.03 (strong endurance types) to 1.10 (speed types who fade). If your half marathon time predicts a faster marathon than your 5K does, you are endurance-leaning; trust the longer race's prediction. Heat, hills, and altitude all break the formula because it assumes similar conditions. And beginners improve faster than the formula allows: a new runner's 5K today predicts a 10K they will beat in two months simply because they are still improving everywhere.
Frequently Asked Questions
How accurate is the Riegel formula?
Quite good for trained runners predicting between 5K and marathon, usually within 3 to 5 percent. It runs optimistic for speedsters moving up to the marathon without endurance base, and for hot or hilly races. Treat it as a trained-potential estimate.
What is the Riegel exponent?
1.06. Predicted time = known time x (new distance / known distance)^1.06. It means pace fades about 8.4 percent each time distance doubles, a pattern Riegel found fitting world records across distances.
Can I predict my marathon time from a 5K?
Roughly. A 25:00 5K predicts about a 3:56 marathon via Riegel, but only with a full marathon buildup. Without the long runs and mileage, expect 10 to 20 minutes slower. The 5K shows your engine; the marathon tests your endurance.
What is better: Riegel or VDOT?
Both work. Riegel is a simple formula you can compute anywhere; VDOT (Jack Daniels) uses tables tuned to training and tends to suit endurance types better. Many coaches cross-check both and take the conservative number.
Why is my predicted marathon time so fast?
Probably insufficient endurance base. Riegel assumes equivalent training for the target distance. If your longest run is 10 miles, the formula 3:5x prediction assumes months of 40-plus mile weeks you have not done.
How recent should the input race be?
Within 4 to 6 weeks, and an all-out effort. Fitness changes, and a heroic race from last year predicts nothing about today. Time trials count if you truly raced them.
Does Riegel work for trail races?
Poorly. Trails add elevation, technical terrain, and altitude that the formula ignores. Predict the road equivalent, then add 10 to 30 percent depending on the course profile.
What distance best predicts the marathon?
The half marathon: it is long enough to test endurance yet short enough to race often. A half predicts the marathon within about 5 percent for trained runners, closer than 5K or 10K inputs.
Should I use Riegel to set my marathon goal pace?
Use it as the optimistic bound and your long-run performance as the realistic bound, then pick something between. Starting a marathon at your Riegel pace without the training to back it is the classic blow-up recipe.
Who was Peter Riegel?
A research engineer at Battelle and a competitive marathoner (2:39) who published the formula in Runner World in 1977 after analyzing world records. It remains the most cited race-prediction equation in running.