We Analyzed 233,917 Workouts: How Hard You Train Affects Your Sleep More Than When

"Evening Workouts Wreck Your Sleep" Is the Wrong Question

You have probably seen the chart. A clean line sloping down, a confident headline, some version of "training in the evening is costing you sleep." It gets posted every few weeks and it always does numbers.

We never quite believed it. The athletes we work with train at every hour you can imagine, and most of them sleep fine. So we went and looked at the data we actually have.

athletedata has 233,917 workouts and 27,609 nights of sleep sitting next to each other, pulled from Garmin, Oura, WHOOP, Apple Watch and the rest, across 745 athletes. That is enough to answer the question properly instead of guessing.

The short version: the hour you train barely moved sleep at all. How hard you trained moved it a lot. And the only place timing actually bit was hard sessions finished close to bedtime.

How We Looked At It (And Why It Matters)

The trick most of these analyses miss is between-person noise. Someone who always trains at 9pm might simply be a worse sleeper than someone who trains at dawn, and that has nothing to do with the workout itself. Compare those two people directly and you will "discover" that evening training hurts sleep, when all you have really found is that night owls sleep worse.

To get around that, we scored every night against that same athlete's own baseline. A value of zero means a normal night for that person. Positive means better than their usual, negative means worse. Then we lined those scores up against the gap between when each athlete finished their last workout and when they went to bed.

This is the same within-person method the serious research uses, and it is the only way to separate "this workout changed my sleep" from "this is just how I sleep."

Finding 1: Intensity Is the Real Driver

Split every workout into easy and hard, relative to that athlete's own normal effort, and the picture separates cleanly.

After a hard session, overnight HRV came in lower and resting heart rate higher than that athlete's usual night. After an easy session, both looked better. The gap between an easy day and a hard day was roughly four times larger than anything timing did on its own.

And it held whether the workout finished at lunchtime or late evening. That is the tell. If the effect followed the clock, an afternoon hard session would look fine and a late one would look bad. Instead, hard days looked worse across the board, because what you are seeing is your nervous system working through the load you gave it. A big training stress shows up overnight no matter when you did it.

If you have ever watched your resting heart rate creep up during a heavy block, this is the same mechanism on a single-night timescale.

Finding 2: Timing Bit in Exactly One Place

There was one corner where the clock mattered, and it was specific: sessions that were both hard and finished within about two hours of bed.

Those were the worst nights in the entire dataset. REM and deep sleep dropped, and HRV went with them. A hard effort that close to lights-out leaves your body still ramped up when it should be winding down, and the first part of the night pays for it.

Give that same session three or four hours of room before bed and most of the penalty went away. So the problem was never "evening." The problem was "hard, and right before sleep."

Finding 3: Easy Sessions Are Fine At Any Hour

This is the part the viral version gets most wrong. An easy spin or an easy jog an hour before bed came out slightly better than that athlete's average night, not worse.

So if your only window to move is the evening, and you keep it easy, you are not sabotaging your recovery. You may even be helping it. Light evening movement and a hard interval session land in completely different places, and lumping them together as "evening exercise" is what produces the scary headline.

This Matches the Research, Not the Clickbait

None of this argues with the science. The largest study on the question (Leota and colleagues, 2025, roughly four million nights of wearable data) found the same shape we did: a timing-by-intensity effect with a rough four-hour line, not a blanket ban on evening exercise. Their numbers put the cost of a very intense late session at up to around 45 minutes of sleep, with light activity barely registering.

The research was never the problem. The problem is what happens when a careful finding gets compressed into a screenshot. "Hard sessions close to bed can cost you some deep sleep" becomes "evening workouts wreck your sleep," and the nuance that would actually help you gets thrown out.

What To Actually Do With This

A few practical rules fall out of the data:

Keep easy work wherever it fits your life. Morning, lunch, late evening. It does not seem to matter, and an easy evening session may even help.
Give hard sessions some daylight before bed. If you can finish your intervals or your heavy lifting three to four hours before sleep, do it. When your schedule does not allow it, accept that the night might be a little lighter and plan recovery accordingly.
Watch your own overnight numbers, not the average. If a late hard session genuinely tanks your HRV, your data will show it. If it does not, stop worrying about a rule that was written for someone else.

That last point is the one we care about most. Everything above is an average of 745 strangers. Useful, but it is not you. Your own couple hundred nights know whether late intervals actually cost you, and by how much. A coach that can read your sleep and training data together can tell you where your line is, instead of handing you a rule of thumb built from other people.

The best athletes do not just train hard. They know when their own body can absorb it.

athletedata connects your training and sleep data and coaches you on what your own numbers actually say. Get started here.