The finding. The "set the treadmill to 1%" rule comes from a 1996 study by Jones and Doust, and its own abstract limits it: nine trained men, runs of about 5 to 6 minutes, and speeds from 2.92 to 5.0 m/s, which is 9:11 down to 5:22 per mile. Nothing slower than 6.5 mph was tested, and at the two slowest speeds they did test, road running was not significantly different from a 0% treadmill either. The ACSM equation answers a different question (what a climb is worth) and says every 1% equals running 4.5% faster on the flat, at any speed.
Where the 1% rule comes from
A. M. Jones and J. H. Doust published "A 1% treadmill grade most accurately reflects the energetic cost of outdoor running" in the Journal of Sports Sciences in 1996. Their starting point, as the abstract puts it, is that running indoors removes air resistance, so a treadmill at a given speed costs less energy than the road at the same speed, and a slight incline can make up the difference.
The design, from the abstract (I read the PubMed record; the full paper is not open access):
- Nine trained male runners, well used to treadmill running.
- Six speeds: 2.92, 3.33, 3.75, 4.17, 4.58 and 5.0 metres per second, each run for 6 minutes with 6 minutes of recovery.
- The whole routine six times: five times on a treadmill (at 0%, 0% again, 1%, 2% and 3%) and once outdoors on a level road.
- Oxygen uptake measured from expired air in the last 2 minutes of each run.
What they found, speed by speed
The abstract reports the result for each speed separately. I converted the speeds to the units runners use (1 mile = 1,609.344 m):
| Speed (m/s) | mph | Pace per mile | Pace per km | Road running matched (no significant difference) |
|---|---|---|---|---|
| 2.92 | 6.53 | 9:11 | 5:42 | 0% or 1% (less than 2% and 3%) |
| 3.33 | 7.45 | 8:03 | 5:00 | 0% or 1% (less than 2% and 3%) |
| 3.75 | 8.39 | 7:09 | 4:27 | 1% only |
| 4.17 | 9.33 | 6:26 | 4:00 | 1% or 2% (more than 0%, less than 3%) |
| 4.58 | 10.25 | 5:51 | 3:38 | 1% or 2% (more than 0%, less than 3%) |
| 5.0 | 11.18 | 5:22 | 3:20 | Between 1% and 2%; not significantly different from any grade |
The authors' conclusion is that a 1% grade matches the energy cost of outdoor running over about 5 minutes at speeds between 2.92 and 5.0 m/s. That is a fair summary, and 1% is the one grade that matched at every speed. But the table shows two things the rule of thumb leaves out. At the slower end of the range, a flat treadmill was also a match. At the fast end, the road sat closer to 2%.
The speeds nobody tested
The slowest speed in the study is 6.53 mph, a 9:11 mile. A lot of treadmill running happens slower than that: 6.0 mph is a 10:00 mile and 5.0 mph a 12:00 mile. The study says nothing about those speeds. What it does show is the direction: the gap between treadmill and road was smallest at its two slowest speeds, where 0% and 1% were both statistically indistinguishable from the road. Air resistance grows with speed, so that direction makes physical sense, but the study does not measure how small the gap becomes below 6.5 mph, and neither do I.
So if your easy runs are at 10:00 per mile or slower, 1% is a harmless habit rather than a measured correction. If you run intervals at 6:00 per mile, the study's own results suggest 1% may slightly under-shoot the road.
What the ACSM equation says an incline is worth
The American College of Sports Medicine's running equation estimates oxygen use (ml per kg per minute) from speed S in metres per minute and grade G as a fraction:
VO2 = 0.2 × S + 0.9 × S × G + 3.5
Set the cost of running at speed S on a grade equal to the cost of a flatter speed S' and solve: 0.2 S + 0.9 S G = 0.2 S', so S' = S × (1 + 4.5 G). At 1% (G = 0.01) the flat-equivalent speed is 4.5% higher. The 0.9 × S × G term is the cost of lifting your body; the equation has no term for air resistance at all. It describes treadmill-style running and is stated to be most accurate above 5 mph (134 m per minute).
Worked example: 6.0 mph is 160.9 m per minute. At 1%, 6.0 × 1.045 = 6.27 mph on the flat, which is 3,600 / 6.27 = 574.2 seconds, or 9:34 per mile instead of 10:00.
What 1, 2, 3 and 5% are worth at common speeds
Each cell is the flat pace per mile with the same oxygen cost in the ACSM equation, and how many seconds per mile quicker that is than the belt speed alone. Computed with the same engine as the treadmill pace chart:
| Belt speed | Pace at 0% | 1% | 2% | 3% | 5% |
|---|---|---|---|---|---|
| 5.0 mph (8.0 km/h) | 12:00 | 11:29 (-31 s) | 11:01 (-59 s) | 10:34 (-86 s) | 9:48 (-132 s) |
| 6.0 mph (9.7 km/h) | 10:00 | 9:34 (-26 s) | 9:10 (-50 s) | 8:49 (-71 s) | 8:10 (-110 s) |
| 6.5 mph (10.5 km/h) | 9:14 | 8:50 (-24 s) | 8:28 (-46 s) | 8:08 (-66 s) | 7:32 (-102 s) |
| 7.0 mph (11.3 km/h) | 8:34 | 8:12 (-22 s) | 7:52 (-42 s) | 7:33 (-61 s) | 7:00 (-94 s) |
| 8.0 mph (12.9 km/h) | 7:30 | 7:11 (-19 s) | 6:53 (-37 s) | 6:36 (-54 s) | 6:07 (-83 s) |
| 9.0 mph (14.5 km/h) | 6:40 | 6:23 (-17 s) | 6:07 (-33 s) | 5:52 (-48 s) | 5:27 (-73 s) |
| 10.0 mph (16.1 km/h) | 6:00 | 5:44 (-16 s) | 5:30 (-30 s) | 5:17 (-43 s) | 4:54 (-66 s) |
In kilometre terms: at 10 km/h (6:00 per km) the flat equivalents are 5:44, 5:30, 5:17 and 4:54 per km for 1, 2, 3 and 5%; at 12 km/h (5:00 per km) they are 4:47, 4:35, 4:24 and 4:05. The pace converter turns any belt speed into pace per mile or per km if your treadmill shows something else.
The percentage gain is fixed (4.5% per 1% of grade), so the saving in seconds per mile shrinks as you speed up: 1% is worth 31 seconds per mile at 5 mph but 16 at 10 mph. Over the study's speed range the ACSM figure for 1% runs from 24 seconds per mile at 9:11 pace to 14 seconds at 5:22 pace.
Two different questions, two different answers
It is easy to mix these up, because both involve "1%":
- Does a treadmill at 1% feel like the road at the same speed? That is the Jones and Doust question. Their answer: at 6.5 to 11.2 mph, roughly yes, for a few minutes, in trained men. Under this view 6.0 mph at 1% is a stand-in for a 10:00 road mile, not a faster one.
- How much harder is a 1% treadmill than a 0% treadmill? That is what the ACSM equation answers: the climbing cost, with no air on either side. Under this view 6.0 mph at 1% costs the same as about 9:34 per mile on a flat belt.
Both can be true at once, because a flat belt is a little easier than a flat road. The treadmill pace chart labels its incline columns as the ACSM flat equivalent for exactly this reason, and treats the Jones and Doust result as a separate question.
Calories on an incline
The same ACSM equation drives the treadmill calorie calculator, at ACSM's 4.9 kcal per litre of oxygen. For a 160 lb (72.6 kg) runner at 6.0 mph for 30 minutes, gross calories come out at 380.7 at 0%, 396.2 at 1%, 411.6 at 2% and 458.0 at 5%. Per mile, a 1% climb adds 0.9 × 0.01 × 1,609.344 = 14.5 ml of oxygen per kg, or 5.2 kcal for that runner, on top of 114.5 net kcal for a flat mile: about 4.5% more, the same ratio as the speed. Speed itself barely changes the cost of a flat mile, which the calories per mile guide works through.
Limits
Jones and Doust tested nine trained men for about 5 minutes per speed; women, recreational runners, long runs and slow speeds were not in the sample. The ACSM equation is an average adult estimate with no air resistance term, and it does not include a downhill saving, so it cannot describe a negative incline. Treadmill displays can also drift from the real belt speed, which the treadmill pace chart shows you how to check. For outdoor hills, where the slope changes and downhills matter, the grade-adjusted pace calculator uses Minetti's measured cost curve instead, explained in the guide to uphill and downhill pace.
Try it: treadmill pace chart, treadmill calorie calculator, grade-adjusted pace calculator.
Sources: Jones AM, Doust JH (1996), A 1% treadmill grade most accurately reflects the energetic cost of outdoor running, J Sports Sci 14:321-327 (PubMed abstract); ACSM's Guidelines for Exercise Testing and Prescription, metabolic equations (Table 7.3 as reproduced by Texas Tech University). Rules checked . Found a problem? Tell us.
Questions
Should I set the treadmill to 1% incline?
It is a reasonable default. Jones and Doust (1996) found a 1% grade matched the energy cost of road running at 6.5 to 11.2 mph over about 5 minutes. At their two slowest speeds a 0% treadmill matched too, and slower speeds were not tested.
How much does a 1% incline add on a treadmill?
In the ACSM running equation, each 1% of grade costs the same as running 4.5% faster on the flat. At 6.0 mph, 1% is worth about 9:34 per mile on the flat instead of 10:00.
Is treadmill running easier than running outside?
At the same speed it costs a little less energy, because there is no air resistance on a treadmill. The 1996 study's results point the same way: at its two slowest speeds a flat treadmill matched the road, and at its faster speeds the road sat between 1% and 2%.
What is 6 mph at a 2% incline equal to outside?
By the ACSM equation, about 9:10 per mile on flat ground (6.54 mph). That counts only the extra climbing, not air resistance.