Pace looks like a single division: take the time you took, divide by the distance you covered, and you have a pace. The arithmetic is easy. The trouble starts one step earlier, because "distance" and "time" rarely arrive in the units you want to train in, and every conversion between them introduces a small error that compounds into a plan built on a wrong number. A runner who wants to know whether a 42.195 km race is realistic at a target finish time, a cyclist planning a long climb, or a walker estimating when they will arrive has to make several unit decisions before the division even happens. Get those decisions right and pace becomes the single most useful training number you own: comparable across days, across routes, and across distances.
The formula
Pace is time per unit of distance. The general form is:
pace = elapsed_time / distance_covered Keep both sides in consistent units. The two common training units are minutes per kilometre (min/km) and minutes per mile (min/mi). If your watch reports elapsed time in seconds and the distance in metres, convert first:
pace_min_per_km = (elapsed_seconds / 60) / (distance_metres / 1000) Speed is the reciprocal, and it is worth knowing because many devices and most training plans speak in speed instead of pace:
speed_km_per_hour = 60 / pace_min_per_km speed_mph = 60 / pace_min_per_mi Converting between the two pace units uses the exact definition of the international mile:
min_per_mile = 1.609344 × min_per_km min_per_km = min_per_mile / 1.609344 A common shortcut is to divide by 1.6 instead of 1.609344. That is fine for mental arithmetic but it inflates the mile pace by roughly 0.6 percent, which is about 24 seconds per mile at a 40-minute pace. Small numbers, but the whole point of pace is that it is comparable.
To turn a pace into a finish prediction for a longer distance, multiply pace by distance and then add nothing else. Runners usually label this "equivalent finish time" rather than prediction, because terrain, weather, and fueling all break the assumption that pace is constant:
finish_time = pace_min_per_km × distance_km Worked example
Suppose a 10 km time trial. The watch records 47 minutes 12 seconds, and the distance is exactly 10.00 km.
Convert the time to minutes: 47 + 12/60 = 47.2 minutes.
pace = 47.2 / 10 = 4.72 min/km That is 4 minutes 43.2 seconds per kilometre, which most runners read as "4:43". In seconds per kilometre it is 283.2 s/km.
Now convert to miles. The trial distance in miles is 10 / 1.609344 = 6.2137 miles, and the pace becomes:
min_per_mile = 4.72 × 1.609344 = 7.596 min/mi So roughly 7:36 per mile. Note that this is not the same as saying "a 7:36 mile pace over 10 km" — the numbers are close because the ratio is fixed, but the habit of thinking in one unit while quoting the other is where errors creep in.
Now project the same effort to a half marathon, 21.0975 km:
finish_time = 4.72 × 21.0975 = 99.58 minutes = 1 h 39 min 35 s And to a marathon, 42.195 km:
finish_time = 4.72 × 42.195 = 199.16 minutes = 3 h 19 min 10 s Now the reverse direction, which is what most race planning actually needs. Suppose the goal is a 3 h 30 min marathon, 210 minutes over 42.195 km:
pace = 210 / 42.195 = 4.977 min/km = 4:58.6 per km Round to a comfortable training pace, say 5:05 per km (5.083 min/km):
finish_time = 5.083 × 42.195 = 214.5 minutes = 3 h 34 min 30 s That gap — about 4.5 minutes — is the distance between the goal and the pace you actually train at. That number, not the pace, is the plan.
Edge cases
- Distance of zero. Pace is undefined; the calculator should refuse or report an error rather than return infinity or zero.
- Negative or zero elapsed time. Physically impossible; treat as invalid input rather than a negative pace.
- Very short distances. A 50 m sprint gives a pace of around 4 s/50 m = 4.7 min/km, which is mathematically fine but measures reaction and flying start as much as fitness. Treat sub-kilometre splits as sprint data, not endurance data.
- Trails and elevation. Pace on hilly terrain is not comparable to pace on flat terrain, even for the same runner. Use gradient-adjusted pace (GAP) if your device supports it, and do not mix GAP with raw pace in the same log.
- Treadmill versus outdoor. Treadmill pace is usually derived from a distance setting that does not match reality; an uncalibrated belt can be off by 2 to 3 percent, which is meaningful over a long session.
- Interval work. Average pace across a session with repeats is almost never the pace you should train at. Read the individual splits instead.
- Multi-sport and mixed sessions. Combining a swim, a run, and a bike into one average pace produces a number that has no physiological meaning.
- Unit mismatch. Entering distance in miles while the calculator assumes kilometres silently shifts the result by a factor of 1.609. Always check the unit labels.
- Rounding display. A watch displaying pace to the nearest 10 seconds per km discards up to 5 seconds of error, which accumulates to more than a minute over a half marathon. Use raw data when precision matters.
Common mistakes
- Dividing distance by time instead of time by distance. The most frequent error. Pace is time per distance; dividing the other way gives speed, and reporting it as pace is off by a factor of distance².
- Using 1.6 for the mile conversion in final planning numbers. It is fine for a rough check, not for a race plan.
- Treating pace as a prediction of the future. Projected finish times assume constant pace and no fade. In practice the second half is slower than the first for most runners, often by 1 to 3 percent.
- Confusing pace with speed in training apps. "5.2" might mean 5:12/km pace or 5.2 km/h speed — a factor of five apart in interpretation. A speed of 5.2 km/h is 11:32 per km, a walking pace.
- Averaging pace across weeks without weighting. A ten-kilometre race pace and a recovery-jog pace do not average meaningfully; the number ends up describing neither effort.
- Ignoring that marathon distance is 42.195 km, not 42. The half marathon is 21.0975 km. Using 21.1 or 42.2 shifts a projected finish by about 10 to 30 seconds, and using the wrong one is a signal that the number was not checked.
- Comparing pace across runners of different abilities. A 5:00/km pace means something entirely different to a first-time marathoner than to a competitive club runner. Compare with yourself, over time.
- Applying race pace to training by default. Training pace is usually slower than race pace by design, often by 30 to 90 seconds per kilometre for most runners.
Try it
The pace calculator at https://www.miryori.org/tools/pace-calculator handles the unit conversions in one place, so you can enter a distance and a time and read the pace directly without doing the division by hand.
YMYL note: pace and projected finish times are estimates based on constant effort and ignore terrain, weather, fatigue, and race-day conditions, so treat them as planning guidance rather than a guarantee.
