Why is my pace slower on hills? How to judge a hilly run
Your pace is slower on hills because every metre of climbing costs extra work, so the same effort covers less ground; judge it by grade-adjusted pace, the pace you'd run on the flat for the same effort, not by the pace on your watch. In the grade-adjusted pace model Pendio uses, a kilometre at +10 % takes 1.55 times as long as on the flat, and at +20 % 2.52 times. This post turns that model into watch paces for each grade, works out why a descent never pays a climb back (at one effort, 1 km up and 1 km down at 10 % take 21 % longer than 2 km of flat), and shows what the model can't see.
TL;DR
- At one effort, Pendio's grade-adjusted pace model has a kilometre take 1.55× as long as on the flat at +10 % and 2.52× at +20 %: an effort worth 6:00 /km on the flat reads 9:18 /km and 15:07 /km on the watch.
- A descent doesn't pay the climb back: at one effort, 1 km up and 1 km down at 10 % take 21 % longer than 2 km of flat, and at 20 % 84 % longer. Beyond about −16 %, a descent is slower than the flat.
- The climb takes most of the clock: on that 10 % hill, 64 % of the time goes on the way up, which drags the average pace to 7:16 /km.
- On inclines steeper than 15.8° (about 28 %), walking cost less energy than running at the same vertical speed in a 2016 study of 15 runners.
- Judge a hilly run by grade-adjusted pace and heart rate, and a climb against your own earlier runs of it. GAP corrects for the grade, not for rocks, mud or the climb you just ran.
How much slower is normal on a hill?
At one effort, a kilometre at +10 % takes about 1.55 times as long as on the flat, and at +20 % about 2.5 times, in the grade-adjusted pace model Pendio uses, a fit to Strava’s heart-rate-based GAP model of 2017 (guide). Here are the model’s time factors (guide) as watch paces, for an example effort worth 6:00 /km (9:39 /mi) on the flat:
| Grade | Time per km against the flat | Watch pace at the same effort |
|---|---|---|
| −30 % | 1.87× | 11:13 /km |
| −20 % | 1.16× | 6:58 /km |
| −10 % | 0.87× | 5:13 /km |
| 0 % | 1.00× | 6:00 /km |
| +10 % | 1.55× | 9:18 /km |
| +20 % | 2.52× | 15:07 /km |
So a kilometre at 9:18 on a 10 % climb and a kilometre at 6:00 on the flat can be the same run at the same effort. Grade-adjusted pace (GAP) puts them on one scale: it’s the pace you’d run on the flat for the same effort, so both read 6:00 /km.
Why does a climb cost so much more than the flat?
On a climb you lift your body as well as carry it forward, so each metre of trail costs more energy the steeper it gets. Minetti and colleagues measured the energy cost of running on a treadmill at 3.40 J/kg/m on the level (joules per kilogram of body weight per metre) and 18.93 J/kg/m at +45 %, more than five times as much (Minetti et al., 2002). A 2017 review sums up the research: the energy cost of running “linearly increases with positive slope” (Vernillo et al., 2017).
Effort is the energy you spend per minute. Hold it steady while each metre costs more and you cover fewer metres a minute: if a kilometre costs 1.5 times the energy, it takes 1.5 times as long.
Runners on real hills meet that limit on the climbs. In a hilly outdoor time trial, oxygen consumption “limited runner’s speeds only on uphill sections”, and the eight experienced runners changed speed mainly through stride length while their stride frequency “was relatively stable” (Townshend et al., 2010).
Why doesn’t the descent give the time back?
A descent saves less time than the same climb costs, and beyond about −16 % it saves none (GPX course analyzer). In the model, a kilometre at −10 % takes 0.87× the flat time against 1.55× at +10 % (guide), and the biggest saving is around −8 %. Strava’s model has the same shape: its help page says the downhill adjustment peaks around −10 % (Strava Support).
Take a made-up hill, 1 km up and 1 km back down at a steady grade, run at the example effort worth 6:00 /km on the flat:
| 2 km of flat | 10 % hill, up and down | 20 % hill, up and down | |
|---|---|---|---|
| Time at one effort | 12:00 | 14:31 | 22:05 |
| Longer than the flat by | – | 21 % | 84 % |
| Average watch pace | 6:00 /km | 7:16 /km | 11:02 /km |
| Share of the time spent climbing | – | 64 % | 68 % |
Two things drag the average down. The descent returns only part of the time: on the 10 % hill the climb costs 3:18 more than a flat kilometre and the descent gives back 47 seconds. And you spend longer on the slow half: almost two thirds of the time goes on the way up, so the average sits much nearer the climb’s pace than the descent’s.
Runners on real hills show the same lopsided pattern. In Townshend’s outdoor time trial, the eight runners ran 23 % slower on the uphills and 13.8 % faster on the downhills than on the level sections (Townshend et al., 2010).
Some road pacing advice says to win the time back on the way down: with a goal of 4:00 /km and a climb at 4:05, “just run 3:55 down to balance it out” (Canadian Running, 2021). On a gentle road hill that can work. On a 10 % trail climb the arithmetic falls apart: an effort worth 4:00 /km on the flat is 6:12 /km up and 3:29 /km down in the model, and balancing the climb would take a 1:48 /km descent, about 33 km/h.
What slows a steep descent isn’t energy. Minetti measured running as cheapest at −20 %, 1.73 J/kg/m against 3.40 on the level, and estimated maximum downhill speeds “well above those attained in downhill competitions” (Minetti et al., 2002). A very steep descent takes longer because “braking costs time too” (guide), and that braking is a load of its own, which training load for trail running counts separately.
Is walking a steep climb a sign you’re too slow?
No: on steep grades, walking can be the cheaper way up. In a 2016 treadmill study, 15 runners walked and ran at seven inclines from 9.4° to 39.2° (about 17–82 %) at a fixed vertical speed of 0.35 m/s, 1,260 m/h. On every incline other than 9.4°, walking cost less energy than running, and the authors concluded that “on inclines steeper than 15.8°, athletes can reduce their energy expenditure by walking rather than running” (Giovanelli et al., 2016). 15.8° is a grade of about 28 %. So on a steep pitch your watch pace drops when you walk, and the energy you spend may drop with it. Whether to walk a given climb in training is for you or your coach to decide.
How should you judge your pace on a hilly run?
Judge a hilly run by grade-adjusted pace and heart rate, and a climb against your own earlier runs of it, rather than by the watch pace alone. Three checks:
- Grade-adjusted pace against your flat pace. If a hilly run’s GAP matches what the same effort gives you on the flat, the hills cost what they should, however slow the watch looked. The Pendio iPhone app shows GAP for every run (guide), and Strava shows its own.
- Heart rate against pace. GAP is a model; heart rate is your body’s reading of the effort. Pendio’s heart-rate cost divides a run’s heart-rate load by its pace load: around your usual, the pace cost what it normally does; higher, it cost more heart rate (guide).
- The same climb, against itself. A climb compared with itself needs no model. VAM, the height you climb per hour in metres per hour, changes with the grade (trail running calculators), so compare it on the same climb. The app recognises a climb every time you run it and lines its efforts up, ranked by score, VAM and GAP (guide).
What can’t grade-adjusted pace see?
Grade-adjusted pace can’t see the ground underfoot or the hill you just ran. Strava says its GAP “does not account for the technical difficulty or condition of the terrain” (Strava Support). A grade-based model reads a rocky, rooty or muddy climb as if it were smooth, so a slower GAP there isn’t a sign of lost fitness.
It also takes each stretch on its own. In the hill time trial, speed “was best predicted using a weighted factor to account for prior and current gradients”: how fast the runners went depended on the hill behind them as well as the ground underfoot. The authors concluded that the level sections after hills offer “the greatest potential to minimize overall time” (Townshend et al., 2010). If your pace sags on the flat after a climb, that is part of the climb’s cost too.
Sources
- Minetti AE, Moia C, Roi GS, Susta D, Ferretti G. “Energy cost of walking and running at extreme uphill and downhill slopes.” Journal of Applied Physiology 93(3): 1039–1046, 2002. doi.org/10.1152/japplphysiol.01177.2001
- Vernillo G, Giandolini M, Edwards WB, Morin JB, Samozino P, Horvais N, Millet GY. “Biomechanics and physiology of uphill and downhill running.” Sports Medicine 47(4): 615–629, 2017. doi.org/10.1007/s40279-016-0605-y
- Townshend AD, Worringham CJ, Stewart IB. “Spontaneous pacing during overground hill running.” Medicine and Science in Sports and Exercise 42(1): 160–169, 2010. doi.org/10.1249/MSS.0b013e3181af21e2
- Giovanelli N, Ortiz ALR, Henninger K, Kram R. “Energetics of vertical kilometer foot races; is steeper cheaper?” Journal of Applied Physiology 120(3): 370–375, 2016. doi.org/10.1152/japplphysiol.00546.2015
- Strava Support. “What is Grade-Adjusted Pace (GAP) on Strava?” Checked 26 October 2026. support.strava.com
- Snider-McGrath B. “Managing effort: learning the art of running uphill.” Canadian Running, 2021. runningmagazine.ca
- Pendio. “How Pendio works”: the training numbers and how split times are worked out. pendio.app/guide
- Pendio. GPX course analyzer: flat-equivalent distance. pendio.app/tools/gpx-course-analyzer
Questions
How much slower should I be running uphill?
At the same effort, a kilometre takes about 1.55 times as long at +10 % and 2.52 times at +20 %, in Pendio's grade-adjusted pace model: an effort worth 5:00 /km on the flat is about 7:45 /km at +10 %. It's a model, not a measurement of you, so compare your own climbs with themselves over time.
Why is my downhill pace barely faster than my flat pace?
Because past a certain grade a descent stops saving time. In Pendio's model descents are quickest around −8 % and slower than the flat beyond about −16 %: a kilometre at −20 % takes 1.16× as long as on the flat, at −30 % 1.87×. Energy isn't the limit there, braking is.
Does grade-adjusted pace account for trail terrain?
No. Strava says its GAP "does not account for the technical difficulty or condition of the terrain". Grade-adjusted pace corrects for the slope, not the surface, so a rocky or muddy climb shows a slower GAP at the same effort.
The bottom line
Expect a climb to cost time and a descent to give back only part of it, and judge the run by grade-adjusted pace rather than by the watch. If a climb still comes out slower than the model, look at the ground, your heart rate and the climb before it, and compare it with your own earlier runs of it. For a race, drop its GPX in the free GPX course analyzer to see the course in flat kilometres at one even effort.