Key Takeaways
- Rolling resistance is Crr × total weight × 9.81 as a force, and that force times speed as watts[1]. An 82 kg rider and bike at 30 km/h on a typical road tire lose about 28 W to it on the flat.
- The tire matters. In Bicycle Rolling Resistance's road table, the tire 10% from the fast end measured 9.7 W on the drum and the tire 10% from the slow end measured 19.0 W, per tire[2][3]. For a rider on two tires at about 29 km/h, that gap is roughly 19 W.
- Those figures come from a drum, which misses the vibration energy that rough roads push into you and the bike[9]. No sourced multiplier exists for each surface, so you can type your own allowance instead.
- With a measured drag area of 0.264 m²[1] and these numbers, air drag takes over from about 17 km/h on the flat. Climbing changes that, because gravity then dominates both.
The Formula
One force, one multiplication, and a grade correction you can usually ignore.
F = Crr × m × g × cos(θ)
P = F × v
Crr is the coefficient of rolling resistance, the ratio of the force holding you back to the load pressing the tire into the road. m is rider plus bike plus anything you carry in kilograms, g is 9.81 m/s², v is speed in metres per second and θ is the angle of the road.
Martin and colleagues wrote it this way in their model of cycling power and assumed Crr does not change with speed[1]. They also point out that cos(θ) is between 0.995 and 1.0 up to a 10% grade, so it hardly matters[1].
Worked example: 82 kg at 30 km/h on the flat
Road, typical gives Crr 0.00417. Force is 0.00417 × 82 × 9.81 = 3.35 N. At 30 km/h (8.33 m/s) that is 27.9 W. In Martin's road trials, rolling resistance made up 10 to 20% of total power, and its share fell as speed rose[1].
Where the Tire Numbers Come From
Fast, typical and slow are places in a lab table, not product names.
Bicycle Rolling Resistance (BRR) runs every tire on a 77 cm drum covered in diamond plate, with a 42.5 kg load at 28.8 km/h in a room held at 21.5 to 22.5 °C. It publishes watts for one tire[2].
BRR notes that an 85 kg rider at that speed loses double the single-tire figure[2]. The watts leave out air drag and bearing drag[6].
To turn those watts into Crr, divide by 42.5 × 9.81 × 8 = 3335.4. That reproduces the Crr values BRR prints in its own tests, for example 15.5 W is 0.0046 and BRR lists 0.0046[8].
For each of BRR's road, gravel and mountain bike tables, the free medium-pressure column is used, tires BRR marks as replaced or archived are dropped, and the tire 10%, 50% and 90% of the way from the fast end is read off.
The tables were read on 3 November 2026, and BRR adds reviews often, so the figures will drift.
| Tire | Tires in table | Drum W / tire | Crr |
|---|---|---|---|
| Road, fast | 156 | 9.7 | 0.0029 |
| Road, typical | 13.9 | 0.0042 | |
| Road, slow | 19.0 | 0.0057 | |
| Gravel / CX, fast | 175 | 15.7 | 0.0047 |
| Gravel / CX, typical | 19.8 | 0.0059 | |
| Gravel / CX, slow | 26.0 | 0.0078 | |
| Mountain bike, fast | 72 | 18.6 | 0.0056 |
| Mountain bike, typical | 24.2 | 0.0073 | |
| Mountain bike, slow | 30.8 | 0.0092 |
Single-tire watts at BRR's medium pressure, 42.5 kg load, 28.8 km/h. Crr is watts divided by 3,335.
"Medium pressure" depends on tire width. For road tires BRR uses about 60 psi (4.1 bar) for a 25 to 27 mm tire[3]. For gravel it is about 42 psi (2.9 bar) for a 39 to 42 mm tire[4], and for mountain bike tires about 29 to 35 psi (2.0 to 2.4 bar) for a 47 to 60 mm tire[5].
The slowest and fastest tires in each table sit well outside these picks: road tires run from 6.7 W (0.0020) to 27.4 W (0.0082). If you know your tire, look it up on BRR, divide its watts by 3,335, and use Own Crr.
The old presets had no source. Road race 0.0033 landed between today's fast and typical road tires, but the mountain bike values of 0.012 to 0.025 sit at or above the slowest mountain bike tire in BRR's table, 0.0122[5].
A drum test cannot say how a downhill tire behaves on a trail.
Why Real Roads Cost More Than the Drum
And why there is an allowance box instead of a surface menu.
A drum test measures what the tire itself wastes as it flexes. On a road, your body and the bike also soak up vibration, and that is lost energy too. Silca calls it impedance and describes total loss as the drum Crr plus impedance.
Below a tire's breakpoint pressure the casing losses dominate, and above it impedance does[9]. Rougher and softer surfaces make impedance climb faster, which lowers the best pressure[9].
In one of Silca's tests, the same road two years later measured a clearly lower Crr than when it was four days old[9].
Rene Herse, a tire maker and so one side of the argument, reports that 42 mm tires rolled equally fast at 30 and 40 psi on real roads with a rider[10].
That fits the idea that wider, softer tires give back on rough ground what a drum says they lose.
Numbers by surface are thin. One field study put a four-wheeled electric bicycle on three surfaces and found Crr of 0.01221 on asphalt, 0.01468 on fine gravel and 0.01832 on coarse gravel[11]. That is about 20% and 50% above asphalt.
It is a different vehicle with small wheels, so the absolute values do not carry over to a bike tire, and two gravel surfaces are not a table. Riders describe a clear drop in speed on chip seal[9], but no trustworthy published ratio exists.
How to use the allowance box
Leave it at 0 on smooth pavement. For rougher ground, add a percentage that matches how it felt, and compare the watts with a ride on your power meter if you have one.
The 20% and 50% above are the only measured gravel-versus-asphalt ratios available, and they come from a different kind of bike.
Speed, Temperature and Pressure
Three things that move Crr a little, with the numbers BRR measured.
Speed. BRR tested six tires at four speeds at one pressure and averaged them. Crr rose about 11.6% while speed went up 150%[6]:
| Speed | Average Crr |
|---|---|
| 14.4 km/h (4 m/s) | 0.00346 |
| 21.6 km/h (6 m/s) | 0.00362 |
| 28.8 km/h (8 m/s) | 0.00378 |
| 36.0 km/h (10 m/s) | 0.00386 |
Average of six tires, 42.5 kg load. BRR, 2022.
This calculator holds Crr fixed at the 28.8 km/h figure, as Martin's model does[1]. Against BRR's averages that reads about 9% high at 14.4 km/h and about 2% low at 36 km/h, which is small next to the choice of tire.
Temperature. BRR heated the tire surface of six tires and measured Crr falling as they warmed[7]:
| Tire surface | Average Crr | Watts / tire |
|---|---|---|
| 20 °C | 0.00398 | 13.3 |
| 30 °C | 0.00357 | 11.9 |
| 40 °C | 0.00328 | 10.9 |
Average of six tires, tire surface temperature, 42.5 kg load, 28.8 km/h.
That is about 10% lower at 30 °C and 18% lower at 40 °C than at 20 °C. This is tire surface temperature, not air temperature. BRR's 22 °C room gives a surface of 24 to 26 °C[7].
BRR did not test below 20 °C, so no sourced figure exists for colder tires. The calculator does not adjust for temperature.
Pressure. On a drum, lower pressure costs watts. One tubeless mountain bike tire measured Crr 0.0046 at 55 psi and 0.0061 at 15 psi[8].
The same test shows the inner tube matters: at 55 psi a butyl tube cost 2.5 W more per tire than tubeless (18.0 W against 15.5 W), and 7.7 W more at 15 psi[8]. On real roads the picture changes past the breakpoint[9][10].
For a pressure to start from, use the Bike Tire Pressure Calculator.
Rolling Resistance and Air Drag
Which one is costing you more at your speed.
Rolling force stays about the same as you speed up, so its power rises in a straight line with speed. Air drag force grows with the square of speed, so its power grows with the cube[1].
That means rolling resistance matters most when you are slow and heavy and least when you are fast and aero.
Example: 82 kg, road typical, 0.264 m² drag area
0.264 m² is the average drag area Martin's riders measured[1]. The calculation assumes air density of 1.2 kg/m³, still air and a flat road. At 20 km/h rolling costs 18.6 W and air 27.2 W.
At 30 km/h rolling costs 27.9 W and air 91.7 W. The two are equal at about 16.6 km/h. An upright rider with a bigger drag area would reach that crossover at a lower speed.
For a complete picture, the Cycling CdA Calculator covers air drag, and the Cycling Power & Speed Calculator puts gravity, air and rolling together. For finish times, try the Cycling Race Predictor.
What It Cannot Tell You
Where this stops being a measurement and starts being an estimate.
- It is not a speed predictor. It gives the watts rolling resistance takes, not how fast you will go.
- The tire tiers are positions in one lab's tables on one day, not your tire. Your own tire could be anywhere in the range.
- One Crr covers both tires. Real bikes carry more weight on the rear wheel, and the two tires may differ.
- Wheel bearing friction and drivetrain loss are separate losses and are not in here[1][6].
- Drum figures are one lab's results. Silca and Rene Herse both sell cycling gear and both say the drum misses real-road effects[9][10].
Frequently Asked Questions
Sources & References
- Validation of a Mathematical Model for Road Cycling Power. Martin JC, Milliken DL, Cobb JE, McFadden KL, Coggan AR. Journal of Applied Biomechanics 14(3):276-291 (1998)
- Our Tests Explained. Bicycle Rolling Resistance. bicyclerollingresistance.com
- Road Bike Tires Test Results (table read 3 November 2026). Bicycle Rolling Resistance. bicyclerollingresistance.com
- CX and Gravel Bike Tires Test Results (table read 3 November 2026). Bicycle Rolling Resistance. bicyclerollingresistance.com
- Mountain Bike Tires Test Results (table read 3 November 2026). Bicycle Rolling Resistance. bicyclerollingresistance.com
- CRR at Different Speeds: Examining Speed Dependence of Rolling Resistance. Bierman J, Bicycle Rolling Resistance. bicyclerollingresistance.com (2022)
- CRR Temperature Test. Bicycle Rolling Resistance. bicyclerollingresistance.com
- TESTED: Rolling Resistance of Tubeless Mountain Bike Tires. Bicycle Rolling Resistance. bicyclerollingresistance.com
- Part 4B: Rolling Resistance and Impedance. SILCA. silca.cc
- Tire Testing: Lab vs Real-Road. Heine J. Rene Herse Cycles (2022)
- What is Slowing Me Down? Estimation of Rolling Resistances During Cycling. Meyer D, Kloss G, Senner V. Procedia Engineering 147:526-531 (2016)

Certified Personal Trainer & Sports Nutritionist
NASM-certified fitness and nutrition coach with over 10 years of hands-on experience helping people build strength, lose fat, and live healthier lives. Specializing in gym-based workouts with a strong focus on lifting technique, biomechanics, and practical exercise science. Through FitLifeRegime, sharing the tools, tips, and insights that have worked for hundreds of clients — helping you start your own fitness journey with confidence and clarity.
Learn more about Manish
NASM Certified
Certified Personal Trainer & Sports Nutritionist from the National Academy of Sports Medicine
Disclaimer: This calculator is for informational and personal journaling purposes only. It logs measurements and compares them to published reference values. It is not health, medical, or nutrition advice.