Lab tested: 40mm road tyres are faster for nearly everyone, and here's why

Over the past 10 years, road tyres have been getting wider, and wider, and wider. From 23mm to 25, then to 28, and now, slowly, the WorldTour is eking its way towards 30mm, while plenty of amateurs have made the jump already.

At Paris-Roubaix this year, some riders will use 35mm tyres which, ironically, would get them kicked out of a cyclocross race, previously given 'wide' tyres to accommodate the rugged courses they face.

But will it ever stop? Or more pertinently, should it? In five years, will the best road bike tyres be wider than today's gravel tyres? Will Tadej Pogačar and Jonas Vingegaard be racing on 40mm road tyres? Or 50!? And should we just cut to the chase and run them now?

Well, Pirelli recently launched the P-Zero Race TLR in a range of widths from 26c to 40c and I saw an opportunity.

It's widely accepted that wider tyres can offer a reduction in rolling resistance. But bigger tyres are, well, bigger, and so aerodynamically they are also theoretically slower.

But does the improvement in rolling resistance outweigh the aerodynamic detriment enough to provide a net gain overall?

Naturally, that question led to others, including:

1. Does a wheel designed around wider tyres - such as a gravel wheel - offset the aerodynamic penalty enough to make it a no-brainer to ride 40mm road tyres everywhere?
2. Given the aerodynamic equation is exponential, at what speed does the aerodynamic penalty outweigh the rolling resistance benefit?
3. How wide is too wide? Is there a tipping point at which wide tyres start to become slower overall? Or in contrast, are we being restricted by bike frame and wheel design? Does the future of aero bikes include 50mm tyre clearance with uber-wide rims?

To try and find some answers, we headed to the Silverstone Sports Engineering Hub and hired out two facilities: the Pedalling Efficiency Rig to test rolling resistance, as with our recent road tyre lab test and gravel tyre test and the Wind Tunnel - the same place we tested 11 superbikes and 18 road wheelsets - to test aerodynamics.

The Pedalling Efficiency Rig is a real-world-applicable testing rig that allows you to test the total wattage loss in a system. By performing multiple tests, keeping everything the same except the tyre, we can confidently conclude that the difference between each test is equal to the difference in rolling resistance between the various tyres.

The Wind Tunnel is a better-known tool in cycling science, but for completeness, it's a system that places an object (in this case, the bike) onto a device that measures drag force. Fans at the rear of the tunnel pull air through an aperture at a controlled speed, pulling air past the bike. The force measured is then used to calculate the object's CdA.

CdA stands for Coefficient of drag x Area, and in Layman's terms, is the measure of how easily air passes around it. In the case of a bike, that then reflects how easily it moves through the air when it's being ridden, so a lower number is good.

The 'Coefficient of Drag' portion is how easily the air passes over the object's surface and is affected by its shape and surface material, while the Area is essentially its frontal size, or how much of it hits the wind.

The tests

To create a fair test, I needed to ensure the tyre on test was the same across all sizes, and with the recent launch of a 40mm road tyre, the obvious choice for the test was Pirelli and its P Zero Race TLR.

This tyre is available in 26c, 28c, 30c 32c, 35c, and 40c with identical construction and casing material. The 'c' in this case is just the standard notation for the width of the tyre, and can be used with 'mm' interchangeably, but as we will see later the actual width of the tyre varies according to the internal width of the rim you are using, so we will stick to 'c' for the most part.

According to Pirelli, they only differ in size. You can feel that the wider tyres are generally thicker in the centre than their narrower counterparts, by dint of simply having more rubber in the mould, but from a construction standpoint, they are as close to equal as we can get, and the 40c is the widest high-end road-focussed tyre that I know exists.

Image 1 of 2

We used two sets of wheels in the test. The 'standard' wheel was the Hunt CGR40. This is a gravel race wheelset that features a relatively wide 25mm internal rim bed, hookless beads, and a 40mm depth. It was chosen partly because we wanted to do some gravel tyre testing on the same day, but also because I wanted our findings to reflect modern standards and help progress them further, rather than testing an already-outdated rim.

Image 1 of 2

For the 'futuristic' wheel, I wanted as wide an internal rim bed as possible, so the natural choice here was the Zipp 303 XPLR SW gravel wheelset, with its 32mm internal rim width and ~40mm outer. These are 54mm deep (bear that in mind when we compare the aerodynamics), with a hookless bead.

We do not recommend you use this combination of wheel and tyre together. The rims are so wide that Zipp has published a list of compatible tyres. The Pirelli tyres are not on that list, so we've only paired them together in a lab testing environment, and not on the open road.

The aim of the test is more to further the development of wider tyres and rims, it is not to promote unsafe wheel-tyre combinations.

We also only tested 32c, 35c and 40c tyres on this rim, because anything narrower seemed unsafe, even in the confines of a lab, and pointless given the theory we were trying to prove.

That led to nine wheel-tyre combinations in total:

• Hunt x 26c
• Hunt x 28c
• Hunt x 30c
• Hunt x 32c
• Hunt x 35c
• Hunt x 40c
• Zipp x 32c
• Zipp x 35c
• Zipp x 40c

The bike used for both tests was a Cube Nuroad C:62 SLT. We opted for a gravel bike because, quite simply, not many road bikes have clearance enough for tyres this wide.

We also paired the rolling resistance portion of this test with our recent gravel bike tyre test and for that, we needed the full 50mm clearance offered by the Cube Nuroad. At under £6k for a SRAM Red XPLR equipped carbon bike with carbon wheels and 50mm tyre clearance, it's also pretty good value in my opinion, but that's by the by.

Rolling resistance: protocol and standardisations

The Pedalling Efficiency Rig puts a rider (in this case, me) on the bike, pedalling on a variety of surface types which are printed onto a gigantic drum. It measures power going into the system via a pair of high-tech power meter pedals, and then measures the power coming out at the drum. The difference between these two figures is the total system loss, including the drivetrain, frame flex and so on.

To explain how this works: The rig adds the appropriate resistance at the drum by reversing the power equation for speed. The engineers enter my CdA, the air density and the speed I want to ride at, and the system calculates the power required (at the drum) to hit that speed, adding the resistance to match. This is similar to erg mode in smart trainers, but at a fixed resistance rather than a reactive one that scales against cadence.

I then pedal the bike at the target speed, and my power through the pedals is forced to increase or decrease depending on the efficiency of the bike (or in this case, the tyre).

We tested each wheel-tyre combination on two different surfaces, at two speeds, and repeated each test, for a total of eight runs per tyre, or 72 runs total.

Each tyre was warmed up by riding for a few minutes, and then tested for 60 seconds per run.

The surfaces we opted for were Smooth Tarmac and Cobbles. There is an in-between option available too but to save time, we opted for the two extremes of the surface roughness spectrum.

Each tyre was tested at 9m/s (32.4km/h or 20.13mph) and 11m/s (39.6km/h or 24.61mph). The speed on the rig is shown in metres per second, and focussing on a round integer seemed easier than trying to match 8.3333 m/s (30km/h) or 11.1111 m/s (40km/h).

We chose these speeds as they are a good reflection of road speeds for the spectrum of keen amateur road cyclists and racers, as well as the speed a which professionals will race over rougher surfaces like cobbles (Gianni Vermeesch's ride at the 2024 Paris Roubaix saw him cover the Arenberg Forest at 42.3km/h, landing in Strava's Top 10).

We tested each tyre set up tubeless with 30ml of Muc-Off sealant, at the recommended tyre pressure suggested by SRAM's tyre pressure calculator. As with our gravel tyre test, we accept that bigger tyres may be given more sealant in a real-world scenario, but our aim here was to isolate the performance of the tyre as much as possible, so we opted for as little as needed to help seal the tyre and no more.

My test speed was generally kept within 0.003 m/s (approximately 0.1kph) but given speed was controlled by my power input, it wasn't metronomic. If at any point I strayed outside of 0.01m/s, the test would be discarded and restarted to ensure accuracy.

The bike was kept in the same gear throughout so that cadence, drivetrain efficiency, any effects of cross-chaining, and my pedalling smoothness were unchanged.

I ensured each test was performed with my hands in the same position, and with as consistent a saddle position as possible to ensure consistent weight distribution on the bike. For extra fairness, I dismounted and remounted the bike after each test to replicate what I'd do between tyre changes.

By repeating each test, we could flag any unusual or anomalous results, repeating again if we felt it necessary. The results are then taken as the average of the two.

Given the rig measures at the rear tyre only, the results are reflective of just one tyre, and since a rider's weight distribution is generally around 55% rear and 45% front, I will multiply the results by 1.818 (1 / 0.55) when calculating the total saving.

Wind tunnel: Protocol and standardisations

For the wind tunnel test, we tested each wheel-tyre combination as a pair.

We tested without a rider present, because like with our recent wind tunnel road wheels test, we could achieve comparable results using less time and more accuracy without a rider. The bike was kept in the same gear throughout the test, and the crank arm was taped to the chainstay to ensure its position remained unchanged. The wheels were spinning to match the wind speed.

We tested at the higher of the two speeds: 11m/s (39.6km/h / 24.61mph). We found during our road wheels wind tunnel test that the aero performance is largely unchanged when switching from 30km/h to 40km/h - what's faster at one speed is generally faster at the other.

When calculating the watts saved, I will solve for both speeds.

We tested at seven yaw angles (the direction at which the wind hits the bike): -15, -10, -5, 0, 5, 10 and 15 degrees.

We tested the Hunt x 30c combo at the start, then repeated it at the end to give us a margin of error of 0.19%. The impact of this will be explained in more detail below.

Prior to each test, the wind tunnel underwent a taring process, similar in principle to the tare on your kitchen scales or zero-offsetting your power meter to ensure accurate readings.

For cleanliness and ease, we used inner tubes, ensuring each had the same length of valve. Tyres were inflated to the pressure recommended by SRAM's tyre pressure calculator, to reflect how much each tyre would expand in real-world use.

Each wheel was checked before going into the tunnel to ensure the tyres were inflated and seated correctly, and they were then mounted into the test bike by a Silverstone Sports Engineering Hub engineer to ensure consistency.

Holding the bike in place was a set of stanchions. No corrections were made for these since they were the same across all tests. We also opted against using disc rotors to speed up wheel changes.

The Results

For the results, I will first share the rolling resistance data to show how much benefit there is for switching to a wider tyre, if any at all.

I will then share the wind tunnel data on its own, to show what the penalty is, if any at all.

And below all that, I will combine the two into a single figure at the two speeds to show the net saving or loss.

I will then dig into these results and offer my thoughts, my conclusions, and any extrapolations I feel we can make, as well as any further questions the conclusions create.

Rolling resistance results

I will start with the narrowest option, the 26c tyre on the Hunt CGR40 rim, and compare the rest to that. This will serve as our benchmark.

Those numbers reflect the total system loss as provided by the Pedalling Efficiency Rig, so includes everything between the pedals and the drum: the tyre, drivetrain friction, frame flex and so on. These are our baseline numbers, which in themselves don't tell us much, but the upcoming comparisons will show us the effect of switching to wider tyres.

As mentioned above, I will multiply the differences by 1.818 to show the effect of two tyres (assuming a 45% to 55% weight distribution).

Negative numbers show a tyre as being faster. A positive number is bad as this means more watts were needed to hold the same speed.

The differences are as follows:

Or to graph that out, here are the differences at 9m/s:

And here are the differences at 11m/s:

Interestingly, on the smooth tarmac surface, the differences are negligible. But as you switch to a rougher surface, wider tyres make a huge difference. On the cobbles, the difference between the 26c and the 40c tyre on the Hunt rim is 29.8 watts at the slower speed, and 65.7 watts at the higher speed.

That grows to 44.7 watts at the slower speed and 80.7 watts at the higher speed when you switch to the wider rim, which balloons the tyre out to 41.5mm. Some of this could be a result of the wheel, but I think it's more likely a result of shaping the tyre better.

I was a little surprised to see no benefit for wider tyres on the smoother surface but given the wider tyre comes with an increase in the rubber thickness, this likely reduces the tyre's ability to deform over the micro-imperfections found on the smooth tarmac surface. I imagine that if the thickness of the 40c tyre was exactly the same as on the 26c, we'd see a marginal improvement here too.

Drawing an imaginary trend line through each of those graphs, we can predict that if the surface was even rougher than cobbles, the trend line would be even steeper, and if the surface was somewhere in between, the trend line would also fall somewhere in between. I'll go into more depth on this later.

Wind tunnel results

Now let's switch to the wind tunnel data.

To start with, here is the yaw angle graph showing the CdA for each tyre on the Hunt wheels.

And here's the same graph for the three tyres on the Zipp wheels.

This data shows that on both rims, the narrower the tyre, the more aerodynamic it is, most likely as a result of having a smaller frontal area, and interestingly, at no point did any of the wheel-tyre combinations 'sail' at higher yaw angles.

During our recent road wheel test, most wheels saw better (lower) CdA figures at 10- and 15 degrees than they did at 5 degrees, creating an M shape in the yaw angle graphs above. Here, though, the shape is a V.

I take that to mean there's an opportunity for designers to further optimise how wide tyres and wide rims interact together with the air, to reduce the CdA, and possibly offset the aero detriment found here.

But now for some comparisons. As above, I'll start with the narrowest option (Hunt x 26c) and compare the rest to that.

For clarity, I have calculated watts required using a weighted CdA (wCdA) using calculations set out by Nathan Barry, 2018.

Our confidence margin for this data is 0.0002M², which equates to 0.2w at 9m/s and 0.37w at 11m/s. To keep the data simple, I will compare the figures directly rather than providing a best- and worst-case calculation.

These 'watts required' numbers reflect the power required to pedal the bike, at the respective speed, at the given CdA. This CdA is for a bike only, so on its own this data isn't very helpful, because bikes can't pedal themselves, but it's useful as a baseline for comparison purposes.

For the table below, negative numbers are good as they indicate a watt saving, whereas a positive number means more watts are needed.

On the Hunt rim, switching from the 26c tyre to the 40c tyre will cost you 5.74 watts at 9m/s, and 10.47 watts at 11m/s.

And at those same speeds, switching from the Hunt wheel with a 26c tyre to the wide Zipp rim with a 40c tyre will only cost you 2.62 watts and 4.78 watts respectively.

We mustn't forget the rim depth difference, but I think it's also important to consider the way the tyre interfaces with the rim, especially when you consider that they differed in rolling resistance too. I added the tyre width and external rim width (both as measured on the day) to highlight this.

Interestingly, the 32c tyre on the Zipp rim is the fastest option overall, albeit within our error margin.

In the graph above, I've calculated the aero saving between the 26c tyre and the 40c tyre (both on the Hunt wheels) at a variety of speeds from 20km/h to 60km/h. This shows that the difference is pretty negligible at lower speeds – just 1.3 watts at 20km/h – climbing exponentially as speed increases.

The combined results

Now let's combine the results to see the total effect of switching to wider tyres.

At both speeds on the smooth tarmac surface, we can see that since the difference in rolling resistance is negligible - if anything it increases slightly with size - and the aerodynamic penalty increases with tyre size, so the total difference is net worse for the wider tyres.

As already mentioned above, if the thickness of the tread didn't scale with size, or if manufacturers are able to design wider tyres with wider rims to 'sail' at higher yaw angles, this result might be reversed, but for the data we have using the wheels and tyres we did, smooth surfaces still favour narrower tyres.

This is especially true at higher speeds since the result of the aero penalty grows exponentially with speed.

Here, it's clear that wider tyres offer a significant benefit. On the Hunt rim, switching from a 26c tyre to a 40c tyre can save you an enormous 55.2 watts at 11m/s. On the wider rim, that saving grows to a frankly ridiculous 75.9 watts!

Looking at the raw data from the Pedalling Efficiency Test, I was riding at 360 watts on the 26mm tyres, a power that I could only hold for around 5 minutes. Switching to the wider tyre would put me down to 284 watts – a power I could hold for well over an hour. I'm not sure who'd want to ride on cobbles for an hour, but that's besides the point.

Of course, 26c is narrow by today's standards, but even compared to 32c, which is the most common choice for riders at Paris-Roubaix, a straight switch to 40c would save 25.8 watts. And if that switch was then paired with a wide rim it grows to 46.5 watts.

Now we just need aero bikes designed with clearances wide enough to handle them...

As a related note, Israel Premier Tech used the Factor Ostro Gravel bikes at Paris-Roubaix in 2024, but they stuck with 32c tyres because anything wider fouled the front derailleur. If they threw caution to the sponsor-correctness wind, they could have saved themselves a lot of watts! Although I'm still not sure they'd have beaten Mathieu van der Poel.

Conclusions

As with most of the tests we do, there is no basic conclusion to our testing, as it will always depend on where you ride, the speed at which you ride, the type of surfaces you cover, whether you're racing or not, and even whether you ride solo or in a group.

But to break it down: On a smooth surface, where there are no imperfections in the road for the tyre to deform around, there's no real benefit for switching to wider tyres, and so the aerodynamic penalty makes them net worse.

But on rough cobblestone roads, the improved rolling resistance of the bigger tyre far outweighs the aerodynamic detriment, even at the higher speed where aerodynamics makes more of a difference.

Real-world surface type

The thing is, not many people ride on smooth tarmac for an entire ride. I know on my rides, I barely get it at all. But rarely do people ride on cobbles either. Most of us are treated to varying combinations of something in between. So how can we use this data to inform our choices?

The graph above shows the wattage difference between the various tyre sizes on the two different surfaces at 9m/s. For a surface that falls somewhere between the smooth and cobbles, it's fair to infer that the available savings would also fall in the middle, following a similar trend line, but at a steepness that scales with how rough the road is.

I reckon a typical British road is around 20% along the spectrum, and so if a 40mm tyre saves 44.7 watts on the Cobbles at 9m/s, we could predict a saving of around 20% of that – 8.94 watts – minus the 2.6w aerodynamic penalty for a net gain of 6.34 watts.

If we were to go for something a little smoother, at say 10% along the spectrum, then that rolling resistance saving halves to 4.47 watts, minus the 2.6w aero penalty for a modest (but still positive) 1.87-watt saving.

Where this gets muddy, however, is that we don't know for definite if the 40mm tyre would continue to be the fastest option on those less-rough surfaces.

Given that rolling resistance is, in Layman's terms, the tyre's inability to deform over imperfections in the road, if said imperfections are only 10mm in height, a 30c tyre might deform around it just as well as a 40c tyre, and thus the savings may hit a plateau as tyres get wider.

As ever, answers lead to more questions, and the only way to answer this for definite would be to return to the Pedalling Efficiency Rig and test on more surface types.

But using the data we have in a real-world scenario, if I were planning to do a gran fondo where I don't know exactly what terrain I am going to encounter, I would happily forego the 2.6-watt aero penalty knowing I could save 44.7 watts if the road gets rough at any point.

Have we found the tyre width limit?

For riding on smooth tarmac surfaces, I think we have. You don't need to run massive tyres to deform around surface imperfections that aren't there.

But for cobbled surfaces (and on rough gravel, since the two are sort of the same thing) I don't think we have.

The savings we found continued to grow from 26c right up to 40c, and I genuinely believe they'd continue growing if we fitted a 45c, 50c or 60c road tyre (if such a thing existed).

The best way to look at it, for me, is that the ideal tyre size scales with the roughness of the surface. Look at the size of the imperfections you're going to have to ride over, and if they dwarf the height of your tyre, you're going to have a rough time.

If your tyre can deform around it without impacting the forward momentum of the solid rim it's mounted to, then you're golden.

Should we all be riding 40mm wide road tyres?

I'm very aware that many of today's best road bikes cannot clear a 40mm wide tyre safely, and so to answer the question posed in the headline of this article... No, but in the future we might.

Given most people ride on a variety of surfaces that can occasionally be very rough, and that the aerodynamic penalty for running a wider tyre is only a couple of watts at the sort of speeds normal people ride at, I think the future of road bike design should offer 40mm of tyre clearance.

But for now, the only type of drop bar bike that can currently run a 40mm road tyre safely is an endurance bike or a gravel bike. In both of those cases, there'd be an aerodynamic penalty from switching away from a modern aero bike.

And so, the sensible approach for most riders right now is to ride the widest road tyre they can safely fit into their bike (and ensure they're inflated to within the ballpark of the ideal pressure, of course).

Based on our data, you might need to put out a couple of extra watts when you're riding at higher speeds on smooth roads (on the front of the bunch), and you might add a few grams to your setup (approximately 159g per tyre between the 26mm and the 40mm) but for the majority of the surfaces you'll ride on, you'll roll that little bit easier, putting less strain on your body and your bike.

You'll also be a little bit safer if you hit anything like a pothole, and you'll probably have more grip in the corners too.

Wider tyres are heavier

This is true, but I'd argue it doesn't matter.

If I assume a rider's CdA is 0.35 using the 26mm tyre, with a total bike+rider weight of 83kg, assuming 2% drivetrain efficiency, you'd need to put out 199.01 watts to hit 20kph on a 3% gradient, or 337.42 watts on a 6% gradient.

Next up, I'll switch to the 40mm tyres and add 318g of mass, but to show the effect of weight alone, I will ignore the aerodynamic and Crr differences. In this scenario, the power required grows by just 0.1 watts on the 3% gradient for a total of 199.11. At the steeper gradient, the power grows by 0.83 watts, to 338.25.

Now, if I increase the CdA to 0.363, reflecting the difference in aerodynamics between the two tyres on the Hunt wheels, the power grows again, by 2.26w, to 201.37w on the shallower slope, and by 1.65 watts to 339.90 on the steeper slope.

Of course, as gradients steepen, the effect of the extra weight grows with it, but unless you're regularly riding up mountains, the effect of the additional weight is going to be negligible for the majority of riding that people do, especially since what goes up invariably comes down.

The total wattage cost there equals 2.36w at a 3% gradient, and 2.48w at a 6% gradient. This isn't nothing, so if your event is likely to be won and lost on a climb, it's worth optimising for, but most of this extra resistance comes via the aerodynamic drag, so weight shouldn't be your primary concern. It's still far exceeded by the potential saving on rougher ground too.

Some may flag the old adage of a gram on the wheel being worth two on the frame, but the science around this is far from clear. The small disparity in weight penalty (far from twofold by any analysis) will be most keenly felt, if at all, under acceleration rather than at steady state.

Is there a speed tipping point?

It's fairly well known that the equation for power (watts) grows exponentially with speed. And so, as speed increases, assuming the rolling resistance remains the same, then there must be a tipping point at which the watts saved are outweighed by the extra air resistance.

If we take the best-case rolling resistance difference on offer here, 80.9 watts, and compare the aerodynamic difference between the same two tyres, you need to be travelling at 22m/s (79km/h) before the aerodynamic penalty is bigger.

But that's not quite the right approach, because the exact wattage saving would also grow, since it is reflective of the coefficient of rolling resistance (CRR), rather than a static watt number. To prove this point, at 9m/s, the difference was not 80.9 watts, but 44.7.

Without an exact CRR figure for each tyre, which isn't given by our system as it measures total system efficiency, we can't calculate the power exactly, but given it's guaranteed to be over 79km/h – a speed nobody is ever likely to ride at except on the steepest of descents, and especially not on cobbles – I don't think I need to.

An unexpected finding

With our Pedalling Efficiency Rig data, we're also able to see that system efficiency actually grows with speed. The faster you go on rough surfaces, the less energy you waste to rolling resistance.

On the smooth tarmac, the efficiency was pretty consistent at around 88%, meaning for every watt I put into the bike, only 0.88 watts were transferred to the road. Crucially, that was about the same across both wheels, on all tyres, and at both speeds.

But on cobbles, at 9m/s, the efficiency ranged from around 63% for the narrower tyre to around 73% for the wider tyre. At 11m/s, that same range was 68% for the narrower tyre to around 78% for the wider tyre, meaning equivalent setups were around 5% more efficient at higher speeds.

This sort of explains why when you ride faster, you can feel as though you're 'floating' over the cobbles, whereas when you ride slowly, it feels like you hit each one.

This tells us that when it comes to riding on cobbles, no matter what tyre you choose to use, you're better off riding harder, as more of your effort will transition to the road.

What is the future of road tyres?

Fifteen years ago, the best road bike wheels were designed to be aerodynamic when fitted with 23c tyres. Five years ago, they were designed around 25c tyres. More recently, they're designed around 28c tyres, and since Tadej Pogačar is running 30c tyres for most of this season, you can be sure more brands will start to optimise around a 30mm width instead.

Tyres are already getting wider, and while there's no real rolling resistance benefit on smooth tarmac, I think there's a strong case for going even wider still, to 35mm, 40mm, and perhaps beyond.

But it relies on brands to make it happen, both in terms of designing wheels to be more aerodynamic with wider tyres, and in terms of bikes being given enough clearance to fit them.

I truly believe there's an opportunity for manufacturers to continually develop their wheels with wider tyres in mind, and if they can cut down that aero penalty - perhaps even remove it entirely - then in three, five or ten years, 40c road tyres could be the no-brainer choice for everyone.

Even now though, I think wider tyres – as wide as you can fit safely into your bike – paired with wider rims, are already the sensible choice for anyone riding on normal roads at normal speeds.

It might not be the right choice for Tadej Pogačar or Jonas Vingegaard, who spend most of their time at high speeds and on smooth tarmac, but it would be my preference as an amateur on typically broken British roads, and indeed most of the rural roads I've sampled across Europe and North America.

And who knows, if Pogačar rides Roubaix this year, there's a potential 46.5 watts on the table over and above a 32c tyre when it comes to the cobbled sectors.

How can I use this information for my own tyre choice?

There's a lot of data here, and while I stand by the conclusion that 40mm tyres would be faster for 'nearly everyone', I'm aware that makes assumptions about what 'nearly everyone' wants, it's a bit catch-all, and that it doesn't apply to everyone's individual needs.

Firstly, if you have a bike that can only accept 30mm tyres, that's your limit. You might be thinking "Josh is just a corporate shill, trying anything he can to get us to buy a new bike!"

But realistically, the implications of buying a new bike go far beyond the tyre choice alone. A new bike that can handle 40mm tyres is, as covered above, likely to be less aerodynamic (in itself, or in the position it puts you) than an aero road race bike. There will be bike fit, weight, and other specification implications that may affect the total system efficiency positively or negatively. In the same way that upgrading the engine on your Peugeot will make it faster, selling your Ferrari to buy the Peugeot and a new engine won't.

With that caveat aside, you need to consider each metric as a sliding scale: The surface you're riding on (and if racing, the surface at which your ride/race might be won or lost); the speed you're riding at and whether or not you're riding in the wind or in a group; and to a smaller degree, the gradients you're riding on.

Put in simple terms, the following rules apply:

• The rougher the surface, the better off you'll be on a wider tyre, with an approximately 80-watt swing between the two extremes (26mm to 40mm) on cobbles.
• The faster you go, the worse off you'll be on a wider tyre, with a 1-watt penalty at 20kph, and a 21-watt penalty at 50kph.
• The steeper the gradient, the worse off you'll be on a wider tyre, albeit with less than 1-watt difference between the two extremes (26mm to 40mm) at gradients of 6%.

With those rules in mind, you can apply your own circumstances and gauge whether wider is actually faster. Our assumption above is that 'nearly everyone' rides on imperfect open roads that flit between smooth and rough, and that the overall gain from the rough surface negates the detriment found on smooth.

If you're a time triallist doing the local Club 10 on a smooth road, you're still better off using narrower tyres. This is because the aerodynamic benefit at your speed (assuming 40-50kph) will far outweigh any rolling resistance difference.

If you're optimising for a hill climb and you know the roads are smooth, then again, the added aerodynamic penalty (even at slower speeds) coupled with the weight gain will also offset any benefits.

If you're an amateur racer looking to eke out a few watts for this Sunday's road race, then you'll need to know the surface type and how you're planning to race. If it's smooth, then as with time trialling above, just stick with what's most aero. Likewise if you're planning on a solo breakaway where you'll be hitting the wind.

However, if the surface is rough, prioritise rolling resistance. If there's a rough patch such as a cobbled section or just a patch of badly-surfaced road, then how pivotal is it? Is it important that you stay with the lead group on this section, or does it fall at an innocuous point on the route where you can lay off and easily catch back on? If it's pivotal, then optimise for that. As ever, answers lead to more questions, and only when you have the full picture can you plan appropriately.

If you're optimising a bike to race Paris-Roubaix, then the potential 80-watt saving you'll get on the cobbles will be very hard to ignore. If you're planning on getting in the early breakaway, which is likely to be ridden at close to 50km/h, then the aerodynamic penalty will likely be too severe to consider it. If, however, you can spend the road sections hiding in the peloton, shielded from the wind, then 40mm tyres would be the way to go. The race is likely 'won and lost' on the cobbles, so why not give yourself every advantage there? Of course, I don't actually think anyone in the WorldTour is reading this looking for bike setup tips, but it's a fun exercise to ponder nonetheless.

The more likely scenario is that someone reading this is prepping for the Paris Roubaix Challenge Sportive, and for those people, where average speeds are slower and the key sections of the event are the roughest sections, the benefit of wider tyres is huge, especially if you hide in the wheels on the road sections.