Running Pace, Speed & Predicted Race Time Calculator

Riegel and the other formulas work best when the reference race is recent and within a factor of two of the target distance; predicting a marathon from a 5K overstates fitness for most runners. Use the longest recent race you have.

The calculator adds about 21 seconds per 10 m of ascent and takes off about 10 seconds per 10 m of descent, a coaching rule of thumb in the range the measured energy cost of running on gradients implies (Minetti 2002), which is why a hilly course with equal ascent and descent is still slower than a flat one.

Yes, in the course adjustment. The file fills total ascent and descent (changes under 1 m are noise and ignored, on a flat course whose elevation spans less than 30 m under 3 m, since a 1 m band adds up a few hundred metres of noise over a marathon; segments steeper than 25% are clamped as data spikes, the way the cycling calculator and the big platforms count), sets the course type to point-to-point with the measured descent and puts the course's mean elevation into the altitude adjustment, switched on from 500 m. A note says when the file's distance differs from the target distance. The course appears on a map with its elevation profile, as on the cycling page: hover or drag over the profile to find the point on the map, open the map large, switch to satellite or 2D. The file is read in your browser and stored only there.

Yes. With a GPX course loaded, pick the race date and start time in the weather adjustment and fetch the forecast: Open-Meteo's hourly forecast for the course midpoint over the predicted run (up to 16 days ahead; recorded weather for past dates, the typical conditions of the same week in earlier years beyond the horizon) fills temperature, humidity, the sun from the cloud cover, the wind and the pressure. The wind is what the heat index (WBGT) instrument sees, about 2 m above ground; without a forecast a light breeze of 2 m/s is assumed. A badge next to the temperature says where the values come from, and it turns to "manual" as soon as you move a slider. The request goes from your browser to Open-Meteo; nothing else is sent.

Above a wet-bulb globe temperature of about 10 °C (13 °C for the elite row, whose times stay flat to about 15 °C WBGT across 1258 elite races, Mantzios 2022) running slows by a set percentage per 5 °C WBGT: 1.5% for the elite and 2.5, 3 and 4% for the other levels (Ely 2007), less for acclimatised runners and more once the race lasts over three and a half hours. The WBGT comes from temperature, humidity, sun and the race's altitude the way weather stations compute it (Liljegren 2008). It is shown with a ±2 °C band and the penalty that band spans, because the sun share, the wind at the instrument and the humidity of the race hour are estimates and WBGT methods differ by 2–3 °C among themselves (Bernard 1999); the prediction uses the midpoint. There is no body-mass field, by design: heavier runners produce more heat per unit of skin and lose more in the heat (Dennis and Noakes 1999; Marino 2000), so a heavy runner sits at the upper edge of their level's penalty; a runner well over 75 kg picks the next lower performance level for the heat.

Daniels-Gilbert, the default, derives your VO₂max from the reference race and is the most robust when the target is within a factor of two of it. Riegel is the simple power law with exponent 1.06 and works well from 5 km to the marathon for trained runners. Cameron was fitted to record performances and also handles 800 m. All three models are always available; with Riegel or Daniels-Gilbert an 800 m personal best is not offered as the reference, so choose Cameron for that.

The model treats 5 to 10 °C wet-bulb globe temperature as optimal, roughly 8 to 14 °C air temperature in the shade, for the elite row up to 13 °C WBGT. Above that, elite runners lose 1.5% per 5 °C and the other levels 2.5 to 4%, because they are out in the heat for longer and often into the warmer part of the day. Below the level's optimum, 4 °C air temperature for the elite to 10 °C for newer runners (El Helou 2012), the cold costs about 1% per 5 °C, capped at 3.5%: a dressed, warmed-up runner loses little down to about −10 °C (Sandsund 2012). Direct sun raises the globe temperature inside the WBGT; the "feels like" figure is shown but does not enter the prediction.

Five to seven days of training in the heat roughly halve the heat penalty, and two weeks or more remove about 75% of it (in an outdoor time trial at 37 °C the power loss went 16, 8 and 3.5% on days 1, 6 and 14, Racinais 2015), provided the heat exposure continues up to race week, since about 2.5% of the adaptation is lost per day without heat (Daanen 2018). Altitude acclimatisation gives less back: aerobic capacity stays reduced, but endurance at race intensity recovers in part, so one to two weeks at altitude take about 30% off the altitude loss and three weeks or more about 50%. Pick the level that matches your preparation and the prediction adjusts.

Up to 1,000 m nothing changes in the model. Above that, aerobic capacity falls by roughly 6.5% per 1,000 m (Wehrlin and Hallén 2006 measured 6.3%), a little more the higher you go, reduced in part by acclimatisation. The calculator takes that loss as the time loss, so a 2,000 m race is predicted several percent slower than the same course at sea level.

Yes. Under the predicted finish time, "Take into race plan" writes the split and the distance into the triathlon planner, which shows where it came from and re-sums the finish time; the nutrition plan takes the splits from there.

Yes. Switch the terrain to Trail, or open the trail calculator at swimbikerun.rocks/trail: the course is run stretch by stretch from the GPX at the energy cost of its gradient, with a terrain class per section read from the map data and your climbing speed (VAM) on the steep climbs; a mountain file switches to the trail by itself. The run after a bike of any length, an extreme triathlon for instance, is planned from the bike's hours and intensity.

Choose the race type and the calculator plans the run leg slower than your standalone prediction: at the planned bike intensity about 2–9% in a sprint, 5–10% over the Olympic distance, 6–12% in a 70.3 and 10–30% in an Ironman®-distance marathon, less for stronger athletes. The elite figures come from paired 10,000 m and half-marathon bests against Olympic and 70.3 run splits, the long-distance figures from a cohort of recreational finishers whose Ironman marathon averaged 1.33 times their marathon best. Set your own bike intensity (IF) against the plan the cycling calculator makes for the distance: 0.05 of FTP over it adds about 3 points on the short distances (measured) and about 8 on the long ones (from the extra glycogen it burns, more than in proportion for larger deviations); riding easier gives the short distances nothing back and the long ones half.