Morning or evening: does the time you train change what you get?

Ask ten cyclists when they train and you will get ten answers, most of them determined by a job, a commute and somebody else's bedtime. Ask them when they should train and the answer is usually a vague sense that the evening is better, or that mornings are for hard people.

There is a real rhythm underneath this. Your body is measurably different at 06:00 than at 18:00, and that difference is large enough to move performance. The question worth asking is not whether it exists, but whether it is big enough to reorganise your week around.

The short answer, stated up front: the time-of-day effect is real and small, and the consistency effect is unglamorous and enormous. If choosing the physiologically better slot costs you one session a week, you have made yourself slower.

The rhythm is mostly temperature

Core temperature follows a daily cycle of roughly 0.5–1.0 °C. It bottoms out in the small hours, usually somewhere between 04:00 and 06:00, climbs through the day, and peaks in the late afternoon or early evening before falling again. Almost everything else that varies across the day in exercise performance — muscle contractile function, joint range of motion, nerve conduction velocity, reaction time — moves roughly in step with it.

The mechanism is unremarkable. Warmer muscle contracts faster and relaxes faster, enzyme kinetics run quicker, and tissue is more compliant. This is the same reason a warm-up works, which is a hint about how to handle a morning session: you are not stuck with the temperature you woke up with, you just have further to climb.

How big is the effect, really

For short maximal efforts — sprints, jumps, peak power — the late-afternoon advantage over early morning is reproducible and typically sits in the low single-digit percentages, occasionally higher. Reviews of short-duration maximal exercise find the pattern consistently, though with plenty of variation between studies and protocols.

For endurance, the picture is muddier. The differences shrink, and they shrink further once you allow a proper warm-up. One study measuring VO₂max repeatedly in athletes found that the variation across a day was more than twice the variation between days — which is a genuinely interesting finding about test scheduling, but it is a statement about a laboratory measurement, not a promise that your Saturday ride will be 5% better after lunch.

Chronotype moves your peak

Group averages hide the thing that matters to you. People differ substantially in the timing of their internal clock, and that timing is measurable rather than a matter of self-image. Roenneberg's large chronotype survey established how wide the distribution is, and how much of it is masked during the working week by alarm clocks.

Applied to athletes, this changes the advice. A study of trained subjects found that diurnal performance variation was better predicted by hours since waking than by clock time, with early types peaking relatively soon after waking and late types a long way further into their day. The spread between an individual's best and worst hours was considerably larger than the tidy group curve suggests.

The practical reading: if you are an early type, a 06:30 ride is much less compromised than the average data implies. If you are a late type, it is worse. Neither fact changes what you should do on a Tuesday, but it does explain why two riders in the same club have opposite opinions about morning training and both are telling the truth.

Does training at your event's time of day matter?

Somewhat, and less than people assume. The circadian rhythm of performance partially adapts to the time you habitually train: reviews of time-of-day-specific training report that improvements tend to be largest when tested at the time of day the training was done. A meta-analysis of resistance training found the same asymmetry for strength — gains showed up at the trained time — while muscle size increased regardless of when the lifting happened.

So the adaptation is at least partly time-stamped. But the size of that advantage is modest, and most of what a dawn start actually costs an amateur is logistical rather than physiological: an unfamiliar breakfast, a rushed warm-up, and the psychological novelty of a full-gas effort at an hour you normally spend asleep. Those are solved by two or three rehearsals, not by moving every session for a season. What to eat before a morning ride covers the breakfast half, and the warm-up matched to the ride covers the rest.

The evening question: will it wreck your sleep?

The received wisdom that evening exercise ruins sleep does not survive contact with the evidence. A systematic review and meta-analysis of evening exercise in healthy people found no harm to sleep on average, and if anything a small increase in slow-wave sleep. The one consistent exception was vigorous exercise finishing shortly before bed — roughly within an hour — which did delay sleep onset and shorten total sleep.

That gives a usable rule. A 19:00 threshold session is fine. A 22:00 threshold session that ends at 23:00 is not, and no amount of adaptation makes it fine, because sleep is doing more for your training than the session was. Sleep is the training session you're not doing makes the case at length.

What to actually do

Two refinements are worth the trouble. First, if you have a genuine choice of slots on different days, put the session that depends on producing high power — VO₂max work, sprints, a test — in the later one, and put endurance riding wherever it fits. That is the only place the effect is large enough to bother exploiting. Second, keep the slot stable rather than optimal. Shuffling between 06:00 on Tuesday and 20:00 on Wednesday asks your body to be ready at both ends of its rhythm and leaves it well prepared for neither.

Winter tends to collapse the choice anyway, because the after-work slot disappears into the dark. That is where an indoor session stops being a compromise and starts being the reason the week happens at all — Moveee Indoor runs in a browser tab, so the barrier between deciding to ride and riding is about thirty seconds. Whatever tool you use, the point stands: the slot you can defend every week beats the slot the physiology prefers.

Build the week around what survives contact with your job, as training around a full-time job sets out, and treat the time-of-day question as a tiebreaker rather than a constraint. The rider who trains at their worst hour, consistently, beats the rider who waits for their best hour and misses it.