Biking usually produces fewer grams of CO2e per kilometer than riding the bus, though the published ranges overlap more than slogans admit. Bike lifecycle figures run 5-80 g CO2e per km. Bus figures run 17.7-101 g CO2 per passenger-km.
Should you always skip the bus for a lower-carbon trip? Occupancy and diet decide that, along with how studies count manufacturing. A single average won't save you.
What a bike emission figure actually includes
A bicycle has no tailpipe. Operational CO2 from the ride itself is zero. Published ranges still fold in the bike itself and the food energy you burn to turn the pedals.
The Stanford course analysis puts conventional biking at 28-80 g CO2-eq per km (0.0278-0.0800 kg CO2-eq/km). BikeRadar reports a lifecycle average near 21 g CO2e per km, with production alone around 5 g per km after spreading an average Dutch commuter bike's 96 kg CO2e manufacturing footprint across the distance it will travel, citing an ECF study. Our World in Data groups most estimates at 16-50 g CO2eq per km.
Thing is, they don't slice the same pie.
Production is a sunk cost. Ride farther and the factory grams per km fall. Food works the other way. Pedal another kilometer and you need more calories. Researchers at ETH Zurich treat unassisted riding as about 25 kCal per km. They note that a regular e-bike can cut that extra food intake by about 46%, and a cargo e-bike by about 70%, because the motor takes a share of the load.
You can spend a long time arguing whether a light aluminum road bike smelted on a coal-heavy grid should count the same as a long-used steel commuter that already paid off its factory emissions years ago, and the spreadsheets rarely agree because lifetime kilometers stay a guess until you actually sell the bike or wear it out.
Diet still moves the food slice. A meat-heavy calorie mix loads more CO2e into each pedal stroke than a plant-heavy one. Rider mass and hills do the rest. Ignore food and manufacturing and every bike looks like zero. Include them and you get that 16-50 g band, sometimes stretching to 80 g.
Why bus emissions run from 17.7 g to 101 g per passenger-km
The Daily Telegraph, citing Sensible Transport, used 17.7 g CO2 per person per km. Cycling UK has published 101 g CO2 per passenger-km. Same mode. Huge spread.
Buses get scored per passenger-km because one engine hauls many people. Fill the seats and each rider's share shrinks. Leave them empty and each remaining rider inherits a fat slice of diesel or electricity.
A study of urban bus lines in Pavia, Italy reports vehicle-km, which is a different unit, but the arithmetic is the same. After a full shift to battery-electric buses, one line was projected at 413 gCO2eq per km of bus travel, about a 68% cut from its diesel baseline. Put 20 people on that bus and each person sits near 21 g. Put eight people on it and each person sits near 52 g. Put two people on a late run and the per-passenger number starts looking like a car.
Headcount changes everything.
Fuel and vehicle size move the starting point too. In that Pavia fleet, 7-meter buses on some lines used about 6-7 MJ/km. The 12-meter buses on other lines used about 15-18 MJ/km. Diesel versus battery and stop-start traffic stack on top before anyone divides by riders.
Peak urban routes with standing loads tend toward the 17.7 g neighborhood. Off-peak and late-night service tend toward the high end. You can't treat one city's packed diesel coach as "the" bus everywhere.
Bike vs bus emissions comparison
| Mode | Range (g CO2e per km or pkm) | Source | What drives it |
|---|---|---|---|
| Bike, production only | 5 | BikeRadar (ECF study) | Factory footprint spread over lifetime km |
| Bike, lifecycle average | 21 | BikeRadar | Production plus use, including food |
| Bike, general range | 16-50 | Our World in Data | Diet, bike type, rider |
| Bike, conventional high | 28-80 | Stanford course | Manufacturing plus food calories |
| Bus, low | 17.7 | Daily Telegraph (Sensible Transport) | High occupancy, efficient operations |
| Bus, high | 101 | Cycling UK | Low occupancy, fuel-intensive running |
The low bike numbers meet the best bus numbers. High bike estimates sit inside ordinary bus service, and to be honest you should read this as overlapping bands, not a winner.
A 10 km urban trip, worked out
Take a flat 10 km hop across town. At BikeRadar's 21 g lifecycle average, the bike costs about 210 g CO2e. At Our World in Data's 16-50 g band, you'll sit somewhere between 160 g and 500 g. Stanford's 28-80 g band is 280-800 g for the same distance.
Now the bus. At 17.7 g per passenger-km, that 10 km costs about 177 g, which can beat a mid-range bike. At 101 g per passenger-km, the bus costs 1,010 g. Almost any bicycle in the table wins that one.
I keep repeating occupancy because it is the whole game on the bus side, it really is the lever that dwarfs bike type and diet once the vehicle is moving, and a crush-loaded articulated bus at rush hour is simply not the same carbon object as a six-person night run even if both get called "the bus." They share a name. They don't share a carbon intensity.
If your 10 km is hilly and you eat a high-carbon diet, slide toward Stanford's upper bound. If you ride an e-bike, the food slice shrinks (ETH's 46% cut on extra intake) while battery manufacturing adds a bit back. The net still usually lands inside the bike ranges above, not the empty-bus range.
How to choose the lower-emission option for one commute
Turns out the practical move is to test the trip you actually take, not the national average.
- Estimate bus occupancy for that time of day. Standing-room peak is the 17.7 g world. A quiet midday or late bus is closer to 101 g. If your transit app shows crowding, use it.
- Pick a bike factor that matches how you ride. Production-only 5 g is too low for a fair fight because it ignores food. 16-50 g is the honest planning band for most people. Use 28-80 g if you want a conservative bike number.
- Multiply each factor by the same door-to-door kilometers. Don't compare a 10 km bike path to an 8 km bus line plus a 2 km walk unless you add the walk, which is operationally near zero.
- Give the bus extra credit only when it's electric and full. The Pavia vehicle-km figure of 413 gCO2eq/km is not a passenger number until you divide by riders.
- For trips under about 10 km on streets you trust, biking wins whenever the bus is thin. When the bus is packed, the two modes can tie, and the bus can win against a high-food, short-lived bike.
Hilliness and weather still decide whether you can actually ride. Carbon isn't the only constraint.
FAQ
Does biking always have lower emissions than the bus? No. Don't count on a slogan. Biking at 16-50 g CO2eq/km can match or exceed a full bus at 17.7 g CO2/pkm. Empty buses are the easy win for the bike.
What makes bus emissions swing from 17.7 g to 101 g per passenger-km? Headcount first. Fuller buses spread fuel and electricity thinner per person. Fuel type and vehicle size add the rest.
Do these bike figures include manufacturing and food calories? Yes. Stanford's 28-80 g/km, BikeRadar's 21 g/km lifecycle and 5 g/km production, and Our World in Data's 16-50 g/km all fold in manufacturing. The higher totals include CO2e from the food energy riders expend.
How should occupancy change my choice? High occupancy pulls the bus toward 17.7 g and closes the gap. Low occupancy pushes it toward 101 g and favors the bike's per-person footprint. Remember that one rule if you remember anything.
If you plan routes in greenmoov.app, line the bike distance up against the same window's bus crowding and pick the lower number for that trip. Write down one week of real commutes. The pattern gets obvious fast.