If you’ve ever spun the cranks on a belt drive bike, you may have noticed they don’t rotate as freely as a chain-driven bike. In other words, there’s extra drivetrain resistance.
So, how much resistance are we actually talking about, and how does it compare to a chain drivetrain across different power outputs?
In this resource, I’ll be answering those questions and much more.
The Advantages and Disadvantages of Belt Drivetrains
Belts offer several advantages over chains, including significantly longer service life (often up to 30,000 km), almost no drivetrain maintenance (no lubes, degreasers, or messy cleanup), near-silent operation, and even a slight weight advantage. They’re especially appealing for bike touring and bikepacking, which helps explain why more than 40 touring bike brands now offer belt-driven bikes in their lineup.
The downsides are that belts require a frame specifically designed around a belt drivetrain, replacement parts can be harder to source, and the upfront cost is usually higher. You can find my full list of pros and cons HERE.
In 2016, Friction Facts published a fascinating comparison between a Gates belt drivetrain and a typical chain drivetrain. The surprising result? Depending on the power output, both systems can come out ahead.
Wait… what?
(You can see the full test report HERE)
Preload Tension
The main source of friction in a chain drivetrain comes from the tension created as the rider pedals. Belt drivetrains work a little differently because their primary source of friction comes from the belt tension that’s built into the system before you even start pedalling.
This high preload tension is necessary on belt drivetrains to prevent the belt from skipping across the sprockets under load. Chains also rely on a small amount of preload tension to stay engaged with the sprockets, but it’s minimal by comparison. In most cases, the weight of the chain itself is enough.
Chain vs Belt Efficiency Test
For the chain drivetrain testing, two premium chain samples were used: a Shimano CN-7901 (Dura-Ace 10-speed) and a SRAM PC-1091R (Red 10-speed). Both chains were tested using an FSA 53-tooth front chainring with a 107 mm effective radius, paired with a SRAM 19-tooth rear cog with a 78 mm effective radius.
For the belt drivetrain testing, two identical Gates CDX belts were used, each measuring 10 mm wide with 125 teeth. These were paired with a Gates CDX 60-tooth front sprocket with a 106 mm effective radius, and a Gates CDX 22-tooth rear cog with a 79 mm effective radius.
Because belts use an 11 mm pitch while chains use a 12.7 mm pitch, different tooth counts were required to achieve nearly identical effective sprocket radii between the two drivetrain systems.
Test Method
In previous chain tests, the Friction Facts test rig was configured to apply different “span tensions” to simulate varying rider power outputs. This made it possible to compare different chains fairly under identical load conditions and isolate frictional differences in a controlled way.
However, because belt drivetrains operate with a much higher inherent preload tension, it isn’t possible to run equivalent span-tension comparisons between belts and chains. As a result, Friction Facts instead measured drivetrain losses using different “total span tensions”, allowing both systems to be evaluated under comparable overall load conditions despite their fundamentally different tension requirements.
Calculating Total Span Tension On A Belt Drivetrain
At a rider output of 250 watts on a belt-drive system, the top span experiences 138.30 lbs of tension. This consists of 53.30 lbs generated by pedal input plus 85 lbs of preload tension already present in the belt. The bottom span experiences 31.70 lbs of tension, which is the result of subtracting the 53.30 lbs of rider input from the 85 lbs of preload (85 − 53.30).
When the top and bottom spans are combined, the total span tension is 170 lbs. In simpler terms, total span tension for a belt drivetrain can be approximated by doubling the span tension.
Calculating Total Span Tension On A Chain Drivetrain
At the same 250-watt rider output on a chain-drive system, the top span is subjected to 54.81 lbs of tension. This includes 52.81 lbs from rider input plus 2 lbs of minimal chain preload. The bottom span experiences approximately 1 lb of tension, largely from the hanging weight of the chain itself.
When these values are summed, the total span tension in the chain drivetrain is 55.81 lbs.
Comparing Frictional Losses Between Drivetrains
Although total span tension differs significantly between belt and chain systems, frictional losses can still be compared directly by matching equivalent rider power outputs on the test curves.
For example, at 250 watts, the belt and chain trend lines can be compared to determine the relative efficiency difference between the two drivetrains.
Additional Test Details
- Belts underwent a 1-hour run-in period at 250 W before testing.
- Chains underwent a 1-hour run-in period at 250 W before testing.
- Both chains were cleaned and re-lubed with a basic light bearing oil before testing.
- Cadence was maintained at 95 RPM.
- Span tension values are measured directly at the span, not derived.
- Full Tension Tester accuracy: ±0.02 watts.
- System losses from ceramic bearings in the test rig were measured and removed from the final results.
Results: Belt vs. Chain Friction
Frictional Losses at 100 watts:
Chain ~1.5 watts
Belt ~2.45 watts
Frictional Difference: ~1 watt (1% advantage to the chain)
Frictional Losses at 150 watts:
Chain ~2 watts
Belt ~2.45 watts
Frictional Difference: ~0.45 watt (0.3% advantage to the chain)
Frictional Loss at 208 watts:
Chain 2.43 watts
Belt ~2.45 watts
Frictional Difference: ~0.02 watt (0.00% advantage to the chain)
Note: One of the most useful aspects of this chain and belt testing is that frictional losses scale almost linearly with rider output (as shown in another test HERE). This linear relationship makes it straightforward to estimate friction at different power levels directly from the graphs. To reflect this, total span tension values have been added at 100 W and 150 W for the chain drivetrain, and at 212 W and 229 W for the belt drivetrain.
Chain vs Belt Efficiency Analysis
It can be a little difficult to interpret the graph at first, so here’s a clearer explanation.
Because of the high preload tension in a belt drivetrain, there is already around 2.45 watts of resistance before any meaningful power is even applied. As a result, a chain drivetrain remains more efficient than a belt until approximately 212 watts of rider output, where the two systems become equal. In practical terms, a chain typically holds a small advantage of around 1 watt at very low power outputs, but this narrows to roughly 0.45 watts at the average power levels of a bike traveller.
Looking back at the original Friction Facts report, most conclusions were drawn using a belt preload tension of 85 lbs. At the time, this was the highest tension recommended by Gates (circa 2010). However, this level is less common today, as modern frame designs and materials have improved.
The lowest currently recommended belt tension from Gates is around 28 lbs (56 lbs on the graph). At this setting, the system produces about 2.45 watts of resistance across power outputs from 0 to 212 watts. This lower tension is typically intended for lighter riders with smooth, consistent pedalling styles, although heavier riders can also use it if the frame is sufficiently stiff, as is the case with the KOGA WorldTraveller setup shown in the accompanying video HERE.
At the other end of the spectrum, the highest recommended belt tension is around 53 lbs (106 lbs on the graph). This is intended for taller, more powerful riders or those with a more dynamic pedalling style. At this setting, resistance increases to roughly 3.1 watts across the 0–229 watt range.
So, if belt drivetrains can be more efficient at higher power outputs, why aren’t they used in Olympic track racing?
- Peak sprint forces would likely require even higher preload tension than tested, which could increase overall friction and potentially negate the efficiency benefit.
- Belt drivetrains make rapid gear ratio changes more difficult compared to chains, which is a major disadvantage in track racing where gearing is frequently adjusted.
Chain Lubricants Are Really Important
Friction Facts has also carried out extensive testing on chain lubricants, and the results show that there can be as much as a 5-watt difference between the best and worst-performing options.
This is significant because any small efficiency advantage a chain drivetrain might have over a belt at low power outputs can easily be erased by using a slow or inefficient lubricant.
The key takeaway is that many common wet chain lubes perform poorly in both efficiency and wear protection. In contrast, wax lubricants tend to deliver both lower friction and better drivetrain longevity. Keeping the drivetrain as clean as possible is also critical for maintaining both efficiency and component life.
If you want a deeper dive into chain lubricant performance, Dave Rome’s excellent article on Velo is well worth a read.
Summary
This lab test found that chain drivetrains are around 0.3% to 1.0% more efficient at low power outputs (below ~212 watts), while belt drivetrains become more efficient at higher power outputs (above ~212 watts).
However, the picture becomes more complex when chain lubrication is taken into account. Differences between chain lubricants alone can amount to roughly 4–5 watts of drivetrain loss, meaning lubricant choice can easily outweigh the small efficiency gap between belts and chains.
It’s also important to note that these results were obtained under controlled, ideal laboratory conditions. In real-world riding, mud, grit, and contamination will play a much larger role in overall efficiency. In particularly dirty conditions, a belt drivetrain may have a practical advantage due to its ability to shed debris more effectively.
Overall, the frictional differences between belt and chain drivetrains are relatively small in practice, and often less important than factors like lubrication, cleanliness, and riding conditions.
Check Out Derailleur Drivetrain Efficiency Testing HERE and Gearbox Testing HERE