How it works
Everything is derived from one number: seconds per kilometre. In Time + distance mode that is your entered time divided by the distance, converted at 1 mile = 1.609344 km; in Pace directly mode you supply it outright. From there the tool prints min/km, min/mi, km/h and mph. Race predictions use Riegel's formula, T₂ = T₁ × (D₂ ÷ D₁)^1.06, applied to 1 mile, 5K, 10K, the half marathon at 21.0975 km and the marathon at 42.195 km.
Riegel scales from whatever distance you entered, which makes that entry the most consequential input on the page — a 5K predicts an optimistic marathon, a recent half predicts a believable one. The four presets load 5K · 25:00, 10K · 50:00, Half · 1:45 and Marathon · 3:45 as time-and-distance pairs in kilometres. Pace directly mode has no distance field and assumes 5 km, so every prediction there extrapolates out from a 5K.
The 1.06 exponent is a population fit, not your physiology. It assumes you have actually trained for the longer distance and says nothing about heat, hills, fuelling or course profile, so marathon predictions from short races habitually read fast. The splits table is always per kilometre even when you enter miles, and it stops at 42 rows, so anything beyond marathon distance gets quietly truncated at that point.
Frequently asked questions
Can I trust a marathon time predicted from a 5K? +
Treat it as a ceiling. The 1.06 exponent assumes you have trained for the longer distance, and extrapolating 5K to marathon crosses a large endurance gap, so the figure usually reads several minutes fast. A recent half marathon gives a far more reliable prediction.
How do I convert 5:00 min/km into minutes per mile? +
Multiply by 1.609344. Five minutes per kilometre is 300 seconds, so 300 × 1.609344 = 482.8 seconds, or 8:03 per mile. Choose Pace directly, enter 5:00 with the unit set to min/km, and both figures appear together.
What exactly is Riegel's formula? +
T₂ = T₁ × (D₂ ÷ D₁)^1.06. It scales a known race time to another distance, with the exponent above 1 capturing that pace slows as distance grows. It was fitted to race results, so it describes trained runners rather than any one individual.
Why are my splits in kilometres when I entered miles? +
The splits table is always per kilometre; the distance unit only controls how your entered distance is converted internally. It generates one row per kilometre up to your distance and caps at 42 rows, so ultramarathon entries stop at the marathon mark.
What time format does the field accept? +
Colon-separated. Three parts read as hours:minutes:seconds, so 0:25:00 is twenty-five minutes. Two parts read as minutes:seconds, and a bare number reads as seconds. The pace field uses the same parser, which is why plain mm:ss works there.
Pace, speed and time are one relationship
Pace is time per distance and speed is distance per time, so they are reciprocals and the tool converts freely between them and total time. The reason pace dominates in running is that it maps directly onto what a watch shows and onto how an effort feels, while speed is more natural for cycling.
The one thing to watch is unit consistency: a pace per kilometre and a pace per mile differ by about 60%, and a target set in one and executed in the other is a substantial error rather than a rounding one.
Why race pace does not scale linearly with distance
Sustainable pace falls as distance rises, and it falls predictably: doubling a race distance typically costs around 5–6% in pace for a trained runner. That is why 5K pace does not hold for a half marathon, and why race-equivalence tables exist.
Course and conditions move it further. Elevation, heat, humidity and wind all cost pace, and heat in particular costs far more than most runners plan for. An even-pace plan on a hilly course is not the same as an even-effort one, and even effort is what usually produces the better time.
