Grade-Adjusted Pace (GAP) Calculator
Find your equivalent effort on any slope, or target a pace at a given gradient. Powered by the Minetti energy cost model.
A GAP (grade-adjusted pace) calculator converts your pace on hills into the equivalent flat-ground effort. TrailMath's free GAP calculator uses the Minetti (2002) energy-cost model to show what your uphill or downhill pace is really worth on the flat.
What is Grade Adjusted Pace?
Grade Adjusted Pace (GAP) converts your actual running pace on a slope to the flat-road equivalent effort. If you run 10:00/km on a 15% climb, GAP translates that to roughly 5:00/km on flat - showing you the true physiological effort behind the slow pace on your GPS watch.
GAP makes it possible to compare effort across completely different terrain. A runner maintaining 6:00/km GAP on a climb is working just as hard as the same runner doing 6:00/km on flat ground - even though their actual pace might show 13:00/km. This is why Strava displays GAP in activity feeds: it reveals effort that raw pace hides.
Why trail runners need GAP more than road runners
Road runners see GAP passively in Strava post-run as a curiosity. Trail runners who race mountain courses need it as a pre-race planning tool.
You cannot use flat pace targets on mountain courses. A runner with a 4:30/km flat pace who targets that pace on UTMB's Col de la Seigne climb will blow up within minutes. The climb demands a pace of 12-15 min/km just to maintain equivalent effort. Without GAP, that runner has no objective way to set a pace target.
Consistent effort - not consistent pace - is the goal in ultras. Heart rate and perceived exertion are the gold standard for effort feedback, but they lag behind changes in terrain and are harder to pre-plan. GAP gives you a predictive effort target you can calculate for each segment before race day.
The tool narrows the gap between Strava users and mountain specialists. Strava owns the GAP metric but not the explanation of how to use it for race planning. Most runners with a 5K road background have no intuition for what their effort "should be" on a 30% climb. GAP gives them an objective anchor.
How to use the GAP Calculator
- Enter your flat reference pace The pace you can sustain on flat ground at your target race effort - easy run pace for an ultra, or threshold pace for a shorter fast race.
- Set the gradient Enter the slope percentage of the segment. Your race GPX file or route profile shows gradient per segment. Positive is uphill, negative is downhill.
- Read the adjusted pace The calculator shows your actual pace target on that gradient to maintain equivalent effort.
- Repeat for each segment Work through each major climb and descent on your course. Group similar gradients to build a practical pacing plan.
- Check the gradient table The full table shows GAP at every slope from -30% to +30%, with power-hike zones highlighted at steep uphill gradients.
The science behind Grade Adjusted Pace
Grade-adjusted pace is grounded in Minetti's cost of transport research - two landmark papers (Minetti et al., 1995 and 2002) that measured the metabolic energy cost of walking and running at gradients from -45% to +45%. The resulting cost curves are the most cited biomechanical basis for GAP calculations used by Strava, Garmin, and this tool.
The cost curve is highly non-linear. Running on flat ground costs approximately 3.4 J/kg/m. At 15% uphill grade, the cost rises to roughly 7.5 J/kg/m - more than double. At 30% grade, it exceeds 13 J/kg/m. This is why a runner "slowing down" on a steep climb may actually be working far harder than on a flat section at twice the pace.
Eccentric braking cost on descents. On gentle downhills (5-15% grade) gravity partially offsets metabolic cost and GAP slows relative to flat. But at steep descents (>20%), the quadriceps must perform large eccentric contractions to absorb kinetic energy and control speed. Minetti's data shows that the metabolic cost of steep descent rises back toward flat-ground levels - and for technical mountain descents with braking, can exceed flat-ground cost. This is the reason your downhill pace in a mountain race is often slower than GAP predicts.
Power hiking crossover. Above approximately 15-20% grade, the biomechanical cost of running exceeds walking at maximum walking speed. Koop (2021) and Minetti both identify this as the gradient at which elite mountain runners transition to power hiking. Running through this threshold wastes energy compared to fast hiking. The gradient table highlights these zones so you can plan hiking sections in advance.
What GAP does not capture. The model assumes smooth, consistent surface. Technical terrain - loose rock, roots, mud, river crossings - adds energy cost through lateral stabilisation, irregular footing, and reduced stride efficiency that is independent of gradient. Trail races on technical singletrack consistently produce actual paces 10-20% slower than GAP at the same gradient. Surface type matters as much as slope angle, and no current model captures it reliably.
Minetti AE et al. (1995). Mechanical determinants of gradient walking energetics in man. J Physiol. 481(Pt 1):235-43. Minetti AE et al. (2002). Energy cost of walking and running at extreme uphill and downhill slopes. J Appl Physiol. 93(3):1039-46. Koop J. (2021). Training Essentials for Ultrarunning. 2nd ed. VeloPress.
Common GAP mistakes in trail race planning
Applying road pace directly to trail courses. The most common and costly mistake. A runner with a 4:45/km flat 10K pace who targets 4:45/km on a trail race with 3000m of gain will bonk on the first climb. GAP is the correction that road runners entering their first trail race most urgently need.
Treating GAP as exact on technical terrain. The Minetti model assumes a smooth running surface. Rocky singletrack, root-covered paths, and muddy descents all add energy cost independent of gradient. Use GAP as a lower bound on technical sections and add a 10-15% buffer to actual pace targets.
Using GAP for grades above 30%. The Minetti cost curve has been measured up to 45%, but the experimental sample size at extreme grades is small and individual variation increases. For sections steeper than 30% (which you will encounter at Hardrock 100 and other mountain races), treat GAP as an approximation and rely more on perceived effort and heart rate.
Not accounting for fatigue in the second half. GAP is calibrated to fresh, rested effort. After 80km of a 100km race, your actual pace at a given effort will be 15-25% slower than GAP predicts. If you use GAP to plan the back half of a long ultra, build in a fatigue multiplier or you will be running harder than you think for each target pace.
Ignoring the power hiking crossover. Runners who try to run every metre of a steep climb often go slower overall than those who power hike at the crossover gradient. If the gradient table shows your section as a hiking zone, planning to run it is not a target - it is a mistake.
GAP in practice at iconic races
UTMB's Col du Bonhomme approach (average 18-22% grade, ~900m D+ over 6km). A runner with a 5:00/km flat easy pace should expect an actual pace of 13-16 min/km through this climb to maintain equivalent effort. Runners who try to push below 10 min/km on this section are well above their sustainable effort ceiling and will pay for it on the descent to Les Chapieux. GAP here: approximately 6:00-7:00/km equivalent effort for a comfortable ultra pace.
Lavaredo's ascent to Rifugio Auronzo (first major climb, technical rocky terrain). The gradient averages 20-25% on the approach to Rifugio Auronzo. At this gradient, power hiking is more efficient than running - the GAP calculator and gradient table both flag this as a hiking zone. Elite runners hike this section deliberately. The technical rocky surface adds another 10-15% to actual time vs smooth-gradient GAP predictions.
Comparing two runners with the same 5km road time. Runner A trains on flat roads and has a 5:00/km flat pace. Runner B trains on trails with similar total volume and the same 5:00/km flat pace. On a mountain race with 4000m of gain, Runner B's hill-trained neuromuscular system and hiking efficiency can produce significantly better results despite identical flat-terrain fitness. GAP reveals when two athletes have the same engine but different gearing for mountain terrain.
Frequently asked questions
Is Strava GAP accurate for trail running?
Strava's Grade Adjusted Pace is a reasonable approximation for smooth, runnable trails but becomes less reliable on technical terrain. Strava uses a proprietary model based on Minetti's cost of transport curves - the same science used here - but applies it uniformly regardless of surface type. A 15% grade on a rocky, rooted singletrack trail costs significantly more energy than 15% on a clean fire road. For race planning on technical courses, use GAP as a starting estimate and add a 10-15% buffer for technical sections.
What is a good grade adjusted pace for trail running?
GAP describes your flat-equivalent effort, so "good" depends entirely on your fitness level. If your flat easy pace is 6:00/km, then a GAP of 6:00/km on a 15% climb means you're running at easy effort even though your actual pace is around 11-12 min/km. For race pacing, target a GAP that matches your goal race effort - typically 5-15% slower than your best flat half-marathon pace for easy/moderate effort ultras, or closer to threshold for shorter fast races.
How do I use GAP to plan trail race pacing?
Choose your target flat-equivalent effort (your GAP target) based on your fitness and race goal. For each major climb in your race, read off the actual pace at that gradient from the calculator. This gives you a pacing plan with consistent effort across different terrain - which is physiologically correct. Running to consistent GAP rather than consistent pace means you go slower on climbs (preserving energy) and can sustain effort on flat and descending sections.
Why is my downhill pace not as fast as GAP predicts?
On gentle downhills (5-15% grade) GAP predicts a pace benefit - gravity assists and the mechanical cost drops slightly. But on steep technical descents (>20%), eccentric muscle loading (quads acting as brakes) increases metabolic cost back above flat-ground levels. Additionally, technical trail surface forces reduced stride length and speed for safety. If your actual downhill pace is slower than GAP predicts, you are likely on steep or technical terrain where the model's smooth-surface assumptions break down.
What is the difference between GAP and effort?
GAP is a normalised pace - it converts your actual pace on a slope to what it would be on flat ground at the same metabolic effort. Effort (often measured by heart rate or perceived exertion) is the direct physiological measure. GAP is a useful proxy for effort because it uses an energy-cost model to make the conversion, but it is not a perfect measure of effort. Individual factors like fitness, fatigue, altitude, and temperature affect actual effort at a given GAP. Use GAP as a planning tool and heart rate or perceived exertion as race-day feedback.
Does GAP account for technical terrain?
No - standard GAP models including this one assume smooth, consistent surface at the given gradient. Technical terrain (loose rock, roots, mud, snow) adds 10-20% to energy cost at the same gradient because of irregular footing, lateral stabilisation effort, and reduced stride efficiency. When using GAP to plan a technical mountain race, treat the calculator output as a lower bound and add a buffer. The more technical your course, the larger the buffer needed.
Want a full training plan built on this science?
TrailMath uses these models to build periodized plans adjusted to your goals and terrain.
Get started free