Let's set apart all pros/cons of chains vs belts, like: ease of maintenance, need of a custom frame, cost, spare part availability, all of that: nope. This question is very focusing only on performance.

I read several reviews, one the most interesting being this one. They measure the performance of the two systems with this tool:

enter image description here

and they come out concluding that:

a conventional chain drive consumes 2.92 watts on average, while the belt eats up 3.93 watts. Although the difference is just 1 watt – not enough for most people to care – this works out as a substantial 34.6 percent.

But unfortunately, also:

That data does have some caveats, though. While the chain drive was tensioned to a typical 2lb, Gates recommends a much higher preload tension of 85lb in order to prevent slippage under load.

So, overall it is to be expected that the old chain is going to be more efficient, according to that. But let's not forget that the belt drive system, though less efficient, is also way lighter!

This also affects performance.

In fact, to this article shows how "the entire belt drive system consisting of a belt and front and rear sprockets weighs nearly half of a standard chain", which is a massive improvement in weight.

enter image description here

So in short: Would the weight gain compensate for the performance loss?

  • 3
    it doesn't sound as though that study tells you anything useful, though. "with belt tension too low to actually ride the bike, the belt is still less efficient". To find anything useful you'd need a study that used realistic tension. You also need to take into account the losses and weight of the hub gear system that is needed with a belt drive. On a track bike, sure, the belt might be lighter, but on the track UCI rules are more important than performance and I don't know whether they allow belts.
    – Móż
    Commented Dec 14, 2015 at 7:47
  • 3
    I don't understand how 1 watt is 34%. A standard bike, loafing along, is demanding on the order of 100 watts from the rider, with peaks up to the 500 watt range. And 200 grams weight is chicken feed as well. The article is meaningless. Commented Dec 14, 2015 at 13:51
  • 1
    @DanielRHicks I think he's saying the difference is 34%, belt vs chain efficiency.
    – ebrohman
    Commented Dec 14, 2015 at 18:25
  • 6
    @ebrohman - Which is a meaningless metric. Commented Dec 14, 2015 at 19:35
  • how does increased tension make the belt drive more efficient? The Gates Belt Drives bikes I've seen all have cogs that the toothed belt fits in, so I don't see how it can slip at all.
    – Johnny
    Commented Dec 29, 2015 at 1:22

1 Answer 1


First things first: A belt is probably slightly less efficient than a properly installed clean chain. The test you link already indicates that. Probably with the tension Gates requires you'll loose a bit more power.

On to your question:

The chain is 200 grams heavier than the belt, of course with the chain you get gears, which you don't get with a belt unless you also install an internally geared hub, which changes the weight of the entire drivetrain heavily the other way, but lets disregard that for now. You loose 200 grams, and you loose 1 watt. From how big a pool do you take that watt? Just to grab something, take a number from here for a cyclist doing a timetrial of 10 miles in 30 minutes: 157 watt.

So this means 1 watt is a loss of about 0.66% for a reasonably fit recreational cyclist.

Given that a reasonably fit recreational cyclist weighs about 80kgs, has a bike of 10kgs, and is carrying 3kg of clothing, gear and filled waterbottles, the 200grams (0.2kg) you save translates to about 0.22% weight loss. So we can already conclude that the weight gain would never compensate for the efficiency loss.

However, when going at any sort of speed, weight is not the factor holding you back, air resistance is. So unless you only bike up the steepest parts of HC climbs in your riding the reality is that the weight loss from a belt drive will never compensate for the efficiency loss.

As far as I can see, the reasons for switching to a belt drive (probably combined with something like a rohloff speedhub) are:

  • Less maintenance
  • Less dirt sticking to all the oily bits (belt stays clean longer)
  • Efficiency degrades less over time between cleaning intervals (meaning a dirty chain is a lot less efficient than a dirty belt)

As a year-round bicycle commuter I can tell you that without a doubt a belt-drive combined with an internally geared hub is way more efficient with my time than a chain could ever be. I loose maybe a few seconds over my 15km (one way) commute (if that) but I gain hours by only having to hose my bike down every other week rather than cleaning and re-oiling my entire drivetrain every few days.

  • Amazing answer. For my purpose, which would be mountain biking and bikepacking, the chain setup looks more interesting then. I care less about the time spent caring of the system and more about the energies I have at the end of an exhausting uphill day.
    – Dakatine
    Commented Dec 14, 2015 at 14:17
  • 4
    I still haven't ever heard any stories about how belt drives perform in the cold. With proper lubrication, a chain becomes only slightly less efficient in the cold. The chain still flexes almost as well. I haven't ever heard that confirmed for a belt drive, and I can't imagine it's true. I imagine a belt drive wouldn't be rideable at say -20F (or colder). Commented Dec 14, 2015 at 15:14
  • 3
    @Dakatine also consider that you can't bodge a belt like you can a chain. Carrying a spare belt negates the weight saving.
    – Criggie
    Commented Dec 14, 2015 at 19:49
  • Re: maintenance penalty. The real interesting comparison would be between a belt drive and a chain drive with chain guard, chain case, full fenders with mud flaps or other types of protection. In other words how effectively can you kill off most of this maintenance time required for exposed chains.
    – Leeroy
    Commented May 7, 2018 at 6:34

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.