Yes, you could ride a motorcycle across a lake, but not on a normal bike at a normal speed. A stock motorcycle on 36 psi tyres has to be doing about 100 km/h before the water's dynamic lift can hold it up, the round rear tyre makes almost no forward thrust in water, and the moment you drop below planing speed the bike sinks like the 180-odd kilograms of metal it is.
With these example inputs (a 250 kg bike and rider, 36 psi tyres, lift coefficient 0.7), calculated in this post from NASA's hydroplaning equation (Horne and Dreher, 1963): 10.35 times the square root of 36 is 62 mph, about 100 km/h.
- About 100 km/h: the planing speed of a 36 psi tyre, from NASA's Horne equation.
- 50–60 km/h is enough once a front ski and paddle tyre multiply the planing area.
- About 5 kW at the water keeps a 250 kg bike on skis planing at 60 km/h.
- About 60 m of glide if the engine cuts, then the bike sinks.
- 8–9 cm of clear lake ice carries a 250 kg bike and rider, by Gold's formula.
Fit a front ski and a paddle-style rear tyre and the speed you need falls to roughly 50 to 60 km/h; that is how a modified ZXMoto MX450 covered a Guinness-verified 40.74 km on open water in October 2026.
This is a tier-3 what-if, so I'll solve it properly:
- why a tyre can float on water at all (NASA worked this out for aircraft landing on wet runways in the 1960s);
- what it takes to stay up, and how much power it needs;
- how far you glide when it goes wrong;
- the other way across a lake, on ice, using the formula Canadian ice-road engineers use.
How can a tyre float on water at all?
On 25 October 2026, Zhang Xue, the founder of the Chinese maker ZXMoto, rode a modified MX450 dirt bike along a river route in his hometown of Mayang Miao, in Hunan province. He did six laps and covered 40.74 km on the water, beating the previous record of 31.8 km. Guinness officials checked GPS, timing and video evidence.
The bike had taken about five months to build: a front ski for steering, a paddle-style rear tyre, sealed components, flotation in case of a fall and remapped power. Until that day, RideApart notes, his longest water ride had been just 20 metres.
RideApart: ZXMoto founder sets Guinness record for riding a motorcycle across water (Oct 2026)A tyre moving fast enough over water is held up by hydrodynamic lift, the same push you feel when your palm skims the surface from a moving boat. Water is dense, about 1,000 kg per cubic metre, so ramming it out of the way generates pressure in proportion to the square of speed.
NASA engineers Walter Horne and Robert Dreher noticed this on a tyre treadmill in about 1957 and formalised it in 1963: at some critical speed the water wedge under the tyre pushes up as hard as the vehicle pushes down, and the tyre leaves the surface.
Their lift equation is the standard aerodynamic one with water in place of air:
where \( F_L \) is lift in newtons, \( \rho \) the fluid density, \( A \) the tyre's footprint area, \( V \) the speed and \( C_L \) a lift coefficient. From test-track data NASA took \( C_L \approx 0.7 \) for tyres. They also noticed that load divided by footprint area is roughly the inflation pressure \( p \), which collapses the whole thing to a famously simple rule:
This is the Horne hydroplaning equation, still used by aviation and road-safety engineers. On a road, it tells you when your tyre stops touching tarmac under a film of water. On a lake there is no tarmac underneath, so it tells you something stronger: the speed below which the tyre cannot hold itself up.
Horne and Dreher pulled together NASA Langley's landing-loads track and treadmill tests on aircraft and car tyres. They define hydroplaning speed as the point where hydrodynamic lift equals the load on the tyre, derive \( V_p = 9\sqrt{p} \) knots from a lift coefficient of about 0.7, and list eight observed signs, from wheel spin-down to loss of directional control. A car tyre at 20 psi carrying 900 lb on 0.5 inch of water was photographed lifting off a glass runway. Hydroplaning needed a minimum water depth of roughly 0.1 to 0.4 inch, depending on tread and surface. The limit for our lake: the equation was built for a thin water film over pavement, not for open water, so I use it here as a lower bound on the problem.
Read the paper (free)How fast does a normal motorcycle need to go to ride on a lake?
About 100 km/h, and that is the optimistic figure. Here is the working for a typical touring bike.
Assumptions. Bike plus rider \( m = 250 \) kg, so weight \( W = mg = 2{,}453 \) N. Both tyres at 2.5 bar, which is 36 psi or 250 kPa. Fresh water, \( \rho = 1{,}000 \) kg/m³. Lift coefficient 0.7 from NASA. I ignore the rider's balance problem for now.
- Footprint area. Load over inflation pressure gives the total contact area of both tyres:
\[ A = \frac{W}{p} = \frac{2{,}453\ \text{N}}{250{,}000\ \text{Pa}} = 0.0098\ \text{m}^2 \approx 98\ \text{cm}^2 \]
Calculated in this post from the example 250 kg mass and 2.5 bar tyre pressure, using NASA's finding that load over footprint area is roughly the tyre inflation pressure.
- Set lift equal to weight and solve for speed.
\[ V = \sqrt{\frac{2W}{\rho\, C_L\, A}} = \sqrt{\frac{2 \times 2{,}453}{1{,}000 \times 0.7 \times 0.0098}} = 26.7\ \text{m/s} \approx 96\ \text{km/h} \]
Calculated in this post from the weight and footprint area above, water density and the NASA lift coefficient of about 0.7.
- Check against Horne's shortcut. \( 10.35\sqrt{36} = 62 \) mph, which is 100 km/h. The two agree, as they should, because the shortcut is this calculation with pressure standing in for load over area.
Then the bad news:
- Entry speed. The bike has to arrive at the shoreline already above 100 km/h, because a tyre that touches water below planing speed digs in rather than climbing on top.
- No thrust. A smooth or knobbly road tyre is a terrible propeller: on tarmac it pushes against a solid; on water it mostly sprays a rooster tail backwards, so thrust falls while drag climbs.
A stock bike launched onto a lake is really a projectile with a slowly dying engine.
See the data as a table
| Hydroplaning speed (km/h) | |
|---|---|
| 22 psi | 78 |
| 25 psi | 83 |
| 29 psi | 90 |
| 33 psi | 96 |
| 36 psi | 100 |
| 42 psi | 108 |
| 50 psi | 118 |
Source: Computed from V = 10.35√p mph, NASA TN D-2056 (1963) · Higher pressure means a smaller, harder footprint, so the tyre needs more speed to be lifted. Valid for smooth or closed-tread tyres.
What is the difference between skimming and planing?
Skimming is a stone bouncing; planing is a boat riding on its own pressure wave, and a motorcycle on a lake has to plane continuously.
Physicist Joseph McCauley's 2018 analysis of boat hulls makes the useful point that lift starts before full planing: as speed rises, a hull climbs its bow wave and makes the jump from being held by buoyancy to being held by hydrodynamic lift. A motorcycle has almost no buoyancy to hold it during that jump. It is too dense to float, so it is either planing or sinking, with nothing in between.
McCauley treats a planing hull as a lifting surface, like a wing or hydrofoil, and builds on his earlier calculation of lift on a submerged delta-wing hydrofoil. He argues, with photographic evidence, that lift begins before planing, that a planing-capable hull makes a sharp transition from buoyancy support to lift while ploughing its bow wave, and that at high speed buoyancy becomes negligible. He derives a lift coefficient for a V-bottom hull that matches empirical data when the wetted area is roughly a triangle near the transom. The limit: it is a single-author preprint aimed at a general physics audience, and it is about boats with buoyancy to spare, which a motorcycle lacks.
Read the paper (free)
How did the record riders actually do it?
They cheated the area term. Since required speed goes as one over the square root of the planing area, multiplying the area by ten cuts the speed by about three. That is what skis do.
| Attempt | When and where | Distance | Bike and changes |
|---|---|---|---|
| Luca Colombo, speed run | Lake Como | 1.9 km at 104 km/h, beating Robbie Maddison's 93.3 km/h | Honda CRF450R with nautical skis and a rear paddle; whole rig 205 kg |
| Luca Colombo, crossing | Sep 2018, Lake Como, Gravedona to Colico | About 5 km in about 5 minutes (~60 km/h average) | RM-Z450, modified for water |
| Previous Guinness mark | Before Oct 2026 | 31.8 km | Modified motorcycle |
| Zhang Xue, ZXMoto founder | 25 Oct 2026, river at Mayang Miao, Hunan | 40.74 km in six laps | ZXMoto MX450 with front ski, rear paddle-style tyre, flotation, sealed engine, remapped power |
Zhang's team spent about five months on the bike. RideApart reports that his first of four attempts was in 2009, and that before the record his longest water ride was just 20 metres. That ratio, 20 m to 40,740 m, is the difference between a bike and a purpose-built planing machine.
A stock bike launched onto a lake is really a projectile with a slowly dying engine.
How much power does a motorcycle need to stay on the water?
Less than you'd guess: my estimate is about 5 kW delivered to the water at 60 km/h on skis, and perhaps 10 to 18 kW at the crankshaft once you account for a paddle tyre's poor efficiency. I use the drag equation from the Savitsky method, the standard way naval architects predict planing-hull resistance, as set out in Alourdas's 2016 SNAME notes.
Assumptions. Same 250 kg. Speed 16.7 m/s (60 km/h). Ski trim angle \( \tau = 5^\circ \). Wetted area of the skis and tyre \( S = 0.25 \) m², mean wetted length 0.5 m. Water kinematic viscosity \( \nu = 1.0\times10^{-6} \) m²/s. The wetted area is my estimate, so I test it below.
- Pressure drag. Water pressure acts at right angles to a flat ski tilted at \( \tau \). Resolve it: the vertical part carries the weight, the horizontal part is drag. So
\[ D_p = W \tan\tau = 2{,}453 \times \tan 5^\circ = 215\ \text{N} \]
- Reynolds number.
\[ Re = \frac{V L}{\nu} = \frac{16.7 \times 0.5}{1.0\times10^{-6}} = 8.3\times10^{6} \]
- Skin friction, ITTC-1957 line.
\[ C_F = \frac{0.075}{(\log_{10} Re - 2)^2} = \frac{0.075}{(6.92 - 2)^2} = 0.0031 \]\[ D_f = \frac{\tfrac12 \rho\, C_F\, S\, V^2}{\cos\tau} = \frac{0.5 \times 1{,}000 \times 0.0031 \times 0.25 \times 16.7^2}{0.996} = 108\ \text{N} \]
Calculated in this post from the ITTC-1957 friction line (Alourdas, 2016) and the assumed speed, length and wetted area above.
- Total drag and power.
\[ D = 215 + 108 = 323\ \text{N}, \qquad P = D V = 323 \times 16.7 = 5.4\ \text{kW} \]
- At the crank. A paddle tyre throwing water backwards is a crude propeller. If it is 30 to 50% efficient (my estimate; marine propellers manage 50 to 70%), the engine must supply 11 to 18 kW plus spray and wave losses.
Two things fall out. The pressure-drag term, \( W\tan\tau \), dominates, which is why water-riders keep the nose low. And the power is modest; the hard parts are thrust and balance, not horsepower.
On a stock bike, a lake crossing is impossible in practice: Horne's equation demands about 100 km/h before the tyres can carry the weight, a round tyre cannot make enough thrust in water to hold that speed, and any slowdown ends in sinking. With a front ski and a paddle rear tyre, the planing area rises roughly tenfold, the required speed falls to about 50 to 60 km/h, and my drag estimate of 323 N needs only about 5 kW at the water. The deciding number is that area: give the bike about 0.25 m² of planing surface and the lake becomes rideable, as the 40.74 km record shows.
What happens if you slow down or the engine cuts?
You get about four seconds. With 323 N of drag on 250 kg, deceleration is \( a = D/m = 1.3 \) m/s². If planing fails somewhere around 40 km/h on skis (an estimate), dropping from 60 to 40 km/h takes
Calculated in this post from the 323 N drag and the example 250 kg mass worked out above; the working follows.
Sixty metres, then the bike stops planing, ploughs and goes under. This is why both record rides were done close to shore with rescue boats, and why Zhang's bike carried flotation.
How sensitive is the answer?
The planing speed moves with the square root of everything, so it is forgiving of small errors but punishing of small areas. Use this to check what you just read: change the mass, the planing area and the lift coefficient.
Minimum speed to plane on water
| If this changes | Then |
|---|---|
| Planing area 98 → 252 cm² | Speed needed falls from 96 to 60 km/h (at \( C_L \) 0.7) |
| Planing area 1,000 cm², but \( C_L \) only 0.3 on shallow skis | About 46 km/h |
| Trim angle 5° → 8° | Pressure drag rises from 215 to 345 N |
| Trim angle 5° → 3° | Pressure drag 129 N, but less lift per unit area, so more area is needed |
| Salt water, density 1,025 kg/m³ | Required speed falls by only about 1% |
| A pillion adds 70 kg | Required speed rises by \( \sqrt{320/250} \), about 13% |
Could you ride across a frozen lake instead?
Yes, and it is the only way most riders ever will. The question is how thick the ice must be, and ice-road engineers answer it with Gold's formula, an empirical fit made in the 1950s and 1960s by Lorne Gold of Canada's National Research Council from records of vehicles and horses that broke through:
with \( P \) the allowable load in kilograms, \( h \) the clear-ice thickness in centimetres and \( A \) an empirical constant in kg/cm². Canadian guidelines use \( A \) between 3.5 and 7 in recent versions. Solve for thickness:
So roughly 8 to 9 cm of good, clear ice for a 250 kg bike and rider, before any safety margin for cracks, currents, slush or white ice, which is weaker. Gold's formula also assumes a moving or short-term load; guidelines cap stationary loads at about two hours because ice creeps and cracks under a parked weight.
Barrette compares 14 ice-road guidelines from Canadian provinces, industry and research bodies. Almost all build on Gold's formula \( P = Ah^2 \), a curve fit to real breakthrough events rather than a theoretical result. Recommended values of A range from 2.5 to 10 kg/cm², and 3.5 to 7 in the most recent documents; one US Army CRREL guide uses 10. The biggest disagreement is how much to discount white ice, which produces large differences in allowed loads. Stationary loads are limited to about two hours because sustained stress causes micro-cracking. The limit: it is a review of rules of thumb for vehicles on freshwater ice, and none of it applies to sea ice or rotten spring ice.
Read the paper (free)
What about riding through water rather than on it?
That is the real-world skill, and it is the opposite technique: slow, steady, in a low gear, with the bike's air intake as the hard limit on depth. The physics of planing says speed keeps you on top; the physics of wading says speed builds a bow wave that can drown the engine.
If you are planning a Himalayan trip, read my guide to water crossings and snow riding technique and the list of the highest lakes you can reach by road in India. For the tyre side, the tyre size markings explainer on RoadHook Machines shows what the numbers on the sidewall mean, including the load index that sets the pressure used above.

Monsoon riders meet the thin-film version of this problem every July: see monsoon onset dates for riders, and remember that the Horne speed for a 29 psi tyre is about 90 km/h on standing water deeper than a few millimetres. More rides and bikes live in the Yatri world.
What's new
- ZXMoto founder Zhang Xue rides a modified MX450 40.74 km over water for a Guinness record RideApart, Oct 2026
- Record run on a river in Hunan beats the previous 31.8 km mark, verified on site by a Guinness adjudicator The Star, Oct 2026
- How the record bike was built: flotation slides under both tyres and waterproofing Visordown, Nov 2026
What if next?
- What if the whole monsoon fell in one hour? The road would turn into a lake 87 cm deep, and planing would be the only way to move.
- What if India's highways were joined into one straight road? It would lap the planet more than three times, or leave it altogether.
- What if you built a road straight up into space? The engine quits long before the tarmac does.
Questions people ask
How fast do you need to go to ride a dirt bike on water?
On ordinary tyres, about 90 to 100 km/h, from NASA's hydroplaning equation at typical pressures, and the bike still cannot make enough thrust to hold it. With skis and a paddle tyre, about 50 to 60 km/h; Luca Colombo's 5 km Lake Como run averaged roughly 60 km/h.
Can a motorcycle float?
No. A motorcycle is mostly steel, aluminium and engine, so its average density is well above water's and it sinks once it stops planing. Record bikes add flotation so that the machine and rider can be recovered.
What is the record for riding a motorcycle on water?
40.74 km, set by ZXMoto founder Zhang Xue on 25 October 2026 on a river in Hunan, China, on a modified MX450 with a front ski and paddle-style rear tyre. It beat the earlier Guinness mark of 31.8 km.
How thick does ice need to be to ride a motorcycle on it?
Gold's formula with A = 4 kg/cm² gives about 8 cm of clear ice for a 250 kg bike and rider. Real guidelines add margins for white ice, cracks and currents, and you should never trust ice you have not measured.
At what speed does a motorcycle aquaplane on a wet road?
On standing water deeper than about 2.5 to 10 mm, NASA's rule gives 10.35 times the square root of the tyre pressure in psi, in mph. At 33 psi that is about 96 km/h. Tread grooves raise the real figure; worn tyres lower it.
Calculated in this post from NASA's rule above and a 33 psi tyre: 10.35 times the square root of 33 is about 59.5 mph, which is 96 km/h.
Sources
- Horne WB, Dreher RC: Phenomena of Pneumatic Tire Hydroplaning, NASA TN D-2056 (1963; lift coefficient 0.7, \( V_p = 9\sqrt{p} \) knots and 10.35√p mph, critical depth 0.1 to 0.4 inch)
- McCauley JL: Hydrodynamic Lift on Boats, arXiv:1808.03313 (2018; lift begins before planing, buoyancy-to-lift transition)
- Alourdas PG: Planing Hull Resistance Calculation, the CAHI Method, SNAME Greek Section (2016; Savitsky drag equation and ITTC-1957 friction line)
- Barrette PD: A review of guidelines on ice roads in Canada: Determination of bearing capacity, NRC and TAC (2015; Gold's formula, A of 3.5 to 7 kg/cm², two-hour stationary limit)
- RideApart: ZXMoto founder sets water-riding record (Oct 2026; 40.74 km, six laps, front ski, paddle tyre, five months' development, longest earlier water ride 20 m, first attempt 2009)
- Visordown: ZXMoto founder sets new world record with 25-mile open water crossing (Nov 2026; previous record 31.8 km, flotation slides)
- The Star: Chinese motorcycle entrepreneur sets world record for riding on water (Oct 2026; Guinness adjudicator on site)
- Visordown: Luca Colombo's Lake Como ride (Sep 2018; about 5 km in about five minutes, RM-Z450)
- webBikeWorld: Luca Colombo beats Aussie water record (104 km/h over 1.9 km on a CRF450R with skis and rear paddle, 205 kg rig; Maddison's 93.3 km/h)
- DirtRider.net forum: Dirt bike riding across water (rider questions on speed and paddle tyres)
