Professional road cyclists can spend hours travelling at around 40km/h, but there is no single average that describes the sport. A mountain stage, a flat sprint stage, a short time-trial and a recovery ride ask different things of a rider. Even within one race, the pace can change dramatically.
The most useful comparison separates whole-race average speed from the pace of a particular climb, an hour of racing or a finishing sprint. It also distinguishes a professional peloton on a controlled race route from a solo training ride with junctions, traffic and stops.
What race speeds actually show
Professional road racing normally takes place with road closures or managed traffic controls, depending on the event and local arrangements. It should not be described as ordinary riding among unrestricted traffic. Riders can use the race environment and shelter in the bunch in ways that are not available on an everyday training ride.
A flat stage does not have to be ridden at maximum pace from the start. If the breakaway is small or the teams are managing fatigue, the peloton may spend much of the day controlling the gap before accelerating for the finish. On other days, repeated attacks or crosswinds make the opening hour exceptionally hard.
Measured 2026 race-speed examples
Averages depend on which elapsed time and distance are counted. For a direct example, official Tour data listed 43.227km/h for Tadej Pogačar’s record-winning 2026 men’s Tour over 3,197km; the 2025 equivalent was 42.728km/h. Within that same 2026 race, the fastest road stage was ridden at approximately 50.9km/h across 161.3km from Vichy to Nevers. These are race statistics, not the solo training speed of a particular rider.
The 2026 Tour de France Femmes covered 1,174.7km, with Demi Vollering’s published winning time of 30h54m51s. Dividing route length by the winner’s recorded classification time gives an indicative 38.0km/h, but bonus seconds and timing conventions mean this should be described as an approximation rather than an exact measured on-road average. Men and women raced different routes and distances: the figures do not form a like-for-like physiological comparison.
| 2026 example | Reported or calculated average | What it represents |
|---|---|---|
| Men’s Tour de France GC winner | 43.227km/h | Official edition average |
| Men’s Tour stage 11 | about 50.9km/h | Fast, relatively flat single stage |
| Tour de France Femmes GC winner | about 38.0km/h | Approximation from published distance and GC time |
Compare these figures with a flat solo training session only after allowing for drafting, road closures, team support, terrain, wind and the different methods used to calculate average speed.
Examples from men’s and women’s racing
| Race example | Recorded speed or time | What it tells you |
|---|---|---|
| 2024 men’s Tour de France, stage 10 | 37.2km/h in the first hour | A flat stage can begin at a controlled pace |
| 2024 men’s Tour de France, stage 13 | 48.2km/h in the first hour | An aggressive opening can be much faster |
| 2024 men’s Tour de France, stage 16 | 37.9km in hour one; 44.8km in hour two | The same stage can change pace substantially |
| 2024 Tour de France Femmes, Rotterdam time-trial | Loes Adegeest exceeded 50km/h; Demi Vollering won in 7min 25.14sec | A short time-trial is a different effort from an all-day road stage |
These are dated examples, not current speed records, and the hourly figures are not whole-stage averages. They illustrate why a single number can conceal the character of a race. The same caution applies when comparing men’s and women’s events: distance, climbing, weather, course design and tactics must be considered alongside the riders.
Why a peloton travels so quickly
Shelter from other riders reduces aerodynamic resistance, allowing the bunch to travel at a speed that would require considerably more work alone. The riders taking turns at the front and those sitting deeper in the group are not necessarily producing the same power, despite moving at almost the same speed.
That does not make riding in the peloton easy. Positioning, accelerations out of corners and closing gaps can be demanding. A modest average power can hide repeated hard efforts, just as a high average speed can hide long sheltered periods.
How fast do professionals train?
There is no meaningful universal training average. A recovery ride is intended to be easy; an endurance ride builds time at a controlled effort; an interval session contains both hard work and recovery. Trying to make all three look fast on a ride-sharing app would undermine the purpose of the easier sessions.
A hilly training ride can have a low average speed while containing very hard climbing. Conversely, a flat group ride with a tailwind may be quick at relatively modest effort. Professionals therefore use measures such as power, heart rate and perceived exertion to guide the session rather than treating speed as the only target.
Average speed is not sprint speed
A finishing sprint lasts seconds, and a time-trial may last minutes or an hour. Neither should be presented as a speed a rider holds throughout a road race. Likewise, a fast downhill section says relatively little about sustainable climbing or flat-road power.
Aerodynamic posture also matters, but the position must remain controllable. Our guide to riding in the drops explains why hand position alone does not determine how aerodynamic a rider is.
Moving time, elapsed time and your own riding
Average speed is distance divided by time, so first check which time is being used. Moving-time averages omit some stops, while elapsed-time averages include them. Automatic pause settings can therefore make two recordings of similar rides look different.
Compare your own rides on similar routes and in similar conditions, and look at the effort as well as the speed. For an event such as Mallorca 312, climbing pace, stops and sustainable fuelling matter more than matching a professional’s flat-road average. The same distinction helps when estimating calories burned while cycling: road speed alone cannot tell you how much work you have done.
