Which marathon distance this page uses
The marathon is defined as 42.195 kilometres, which is 26.2188 miles. The familiar "26.2" is that number rounded to one decimal, and the rounding is not free. At 4:30 per mile the two definitions give 1:57:59 and 1:57:54 respectively, five seconds apart. Going the other way, a 3:30:00 marathon works out to 8:00.6 per mile against the official distance and 8:00.9 against the rounded one. Everything on this page uses the official 42.195 km, so mile paces are computed against 26.2188 miles and half marathon figures against 13.1094. If you are pacing off a watch that rounds, expect a handful of seconds of disagreement over a full marathon and none of it to matter.
The other race distances are exact by definition where they are metric: a 5K is 5,000 m or 3.1069 miles, a 10K is 6.2137 miles. Where they are imperial they are exact too, because the international mile has been 1,609.344 metres exactly since 1959. There is no approximation anywhere in the conversion; the only approximation is in the prediction.
Reading a pace against a treadmill
Pace and speed are reciprocals, which makes some pairs worth memorising. Six minutes per mile is 10.00 mph. Ten minutes per mile is 6.00 mph. Five minutes per kilometre is exactly 12.00 km/h and six minutes per kilometre is exactly 10.00 km/h. Those four anchors let you estimate anything in between without arithmetic.
The number on the console is not the number outdoors, and the difference runs in both directions. A moving belt removes air resistance and returns some energy at footstrike, which makes a given speed marginally easier than the same speed on a road; the common correction is a 1 percent grade, which comes from a small 1996 study of trained runners at speeds above about 7 mph and does not transfer cleanly to slower paces. Working against that, belts drift out of calibration, particularly under a heavier runner, and a machine reading 7.5 mph may be doing something else entirely. Compare treadmill sessions to other treadmill sessions on the same machine and treat cross-comparisons as rough.
What the Riegel column is doing
Pete Riegel published the formula in 1977 in a running magazine and refined it in a 1981 paper: predicted time equals known time multiplied by the distance ratio raised to the power 1.06. The exponent is the whole idea. If it were 1.00, doubling the distance would double the time and pace would be distance-independent, which is plainly false. At 1.06, doubling the distance multiplies time by 21.06, about 2.085, so pace decays roughly 8 to 9 percent per doubling.
Riegel fitted that exponent to competitive race results across distances from 3 km to the marathon. Three limits follow directly. It assumes you are racing both distances, not jogging one of them, so predicting a marathon from a training run gives nonsense. It assumes your endurance base actually supports the target distance, which is why 5K-to-marathon predictions run optimistic for people whose long run is eight miles; the honest fix is to predict from the longest race you have genuinely completed. And it degrades outside its fitted range, so predicting a 100-mile finish from a 5K is not a thing the formula was ever asked to do.
| Known result | Riegel marathon prediction | Implied marathon pace |
|---|---|---|
| 5K in 25:00 | 4:00:41 | 9:11 /mi |
| 10K in 50:00 | 3:50:11 | 8:47 /mi |
| 10K in 42:00 | 3:13:22 | 7:22 /mi |
| Half in 1:45:00 | 3:38:41 | 8:20 /mi |
| Half in 1:30:00 | 3:07:26 | 7:09 /mi |
Even splits, and why the table shows them anyway
The split table divides the distance evenly, which is a reference line rather than a race plan. The observed pattern in marathon field data is that most finishers run positive splits — second half slower — and that the slowdown correlates with how far ahead of goal pace the first half was. Elite races and personal bests are disproportionately negative or even split. That association is not proof that going out slow causes the faster finish, since the same fitness that lets you hold pace also lets you judge it, but the practical advice survives either reading: the first three miles feeling too easy is the correct sensation.
Course profile breaks even splits on purpose. On anything with sustained climbing, holding pace up a grade costs disproportionately more energy than the equivalent time saved coming down, so effort-even and pace-even are different plans and effort is the one that finishes. Use the split table on a flat course and use it as a rough checkpoint on anything else.
The three-field solve
Fill any two fields and the third is arithmetic. Fill all three and distance and time win, and the pace box is recalculated from them rather than trusted — if what comes back disagrees with what you typed, the typed pace was wrong or the distance was measured by GPS through a tree line. Time accepts three formats: a bare number is minutes, one colon is minutes and seconds, two colons is hours, minutes and seconds. Pace is always minutes and seconds with one colon.
Questions people ask
Why does my watch report a different pace for the same run?
Almost always distance, not time. Wrist GPS accumulates error on tree-lined trails, in street canyons and around switchbacks, and it consistently reports slightly long on courses with a lot of turning because it cuts corners into straight segments while you run the arc. On a certified road course the course is the truth and your watch is not; certified courses are measured by calibrated bicycle along the shortest legal running line, which is shorter than the line you actually ran through the field. Two to three percent long on a watch over a marathon is unremarkable, and that is about six to eight minutes of apparent pace difference.
Should I train at the pace this calculator gives?
Not for most of your mileage. The consistent finding across endurance training research is that the majority of volume should sit at an easy, conversational intensity, with a small fraction at genuinely hard efforts. Goal race pace is neither. Running every session at the pace you want to race is the most common way recreational runners plateau, because it is hard enough to accumulate fatigue and not hard enough to drive adaptation. Use the goal pace for specific race-pace sessions and for pacing the race itself.
How do I convert a mile pace to a kilometre pace in my head?
Divide by 1.6, or take the mile pace, subtract a third of it, then add a twentieth back — but a lookup is faster. Common anchors: 6:00/mi is 3:44/km, 7:00/mi is 4:21/km, 8:00/mi is 4:58/km, 9:00/mi is 5:36/km, 10:00/mi is 6:13/km. The exact factor is 1.609344, so a mile pace is always the larger number.
What pace do I need for a specific marathon finish time?
Against the official 42.195 km: 4:00:00 is 9:09.5 per mile, 3:45:00 is 8:35.2, 3:30:00 is 8:00.6, 3:15:00 is 7:26.3, and 3:00:00 is 6:52.0. Set the distance selector to Marathon, type the goal time and leave pace blank to get any other target. Add a little for the distance you will actually cover — on a crowded course, running the tangents perfectly is not possible, and most finishers cover a hundred metres or more beyond the measured line.