What's the best time of day to run?
The best time of day to run is whichever slot actually survives a normal week. Tracking 375 members of the National Weight Control Registry, Leah Schumacher and colleagues found exercisers who trained at a consistent time of day, within any of four broad bands, logged more active days and hit the 150-minutes-a-week activity guideline more often than exercisers whose timing varied (Schumacher et al., 2019, Obesity). A separate, much smaller 2025 laboratory study of 12 men found 10-km time-trial performance followed a measurable daily rhythm, fastest around 6:44 p.m. in that sample (Edwards et al., 2025, Chronobiology International). The first study's sample is more than 30 times the size of the second's.
How time of day compares
| Time of day | What the evidence says | Practical trade-off |
|---|---|---|
| Early morning (roughly 4:00–9:00 a.m.) | The largest single group, 47.8%, among consistently-timed exercisers in Schumacher et al.'s 2019 National Weight Control Registry sample trained in this window; the same paper found no significant difference in activity levels between specific time bands (p = 0.06–0.91 across its three measures). Cleveland Clinic cardiologist Erik Van Iterson points out that morning exercise may be associated with lower blood pressure and better sleep. | Runs before the day's obligations can crowd it out. The Cleveland Clinic article states that body temperature and flexibility, tied to aerobic performance, tend to peak later, in the afternoon or evening. |
| Midday (roughly 10:00 a.m.–2:00 p.m.) | Two of Edwards et al.'s six laboratory test times, 10:00 and 14:00, fell in this window; the 2025 study reports only the overall fitted 24-hour rhythm (p < 0.05); it does not report a separate result for each individual clock time. | Falls inside the standard 9-to-5 workday, so it depends on a flexible job, a lunch break or a school schedule that allows it. Also the least distinct evidence of the four rows here, since neither source singles this window out on its own. |
| Early evening (roughly 5:00–8:00 p.m.) | The acrophase, or fitted peak, of Edwards et al.'s 24-hour rhythm in 10-km finish time landed at 18:44, with about 5.0% variation across the day (d = 0.80) in their 12-runner sample. The Cleveland Clinic article states afternoon and evening are when body temperature and flexibility tend to peak. | Matches the fastest average window in that one small, all-male study. Also a window that competes for the same block of time as evening plans, dinner or end-of-day fatigue. |
| Late night (after roughly 9:00 p.m.) or overnight for shift workers | None of the sources above tested this window directly. Edwards et al.'s latest clock time was 22:00, one point inside their six-point rhythm. It stands alone; the study includes no dedicated late-night comparison. | Whatever slot ends up being "late" on an individual's schedule, the consistency finding from Schumacher et al. above still applies to keeping it steady week to week. |
This table is Runked's own, built by lining up the two cited studies' and the Cleveland Clinic commentary's findings by time-of-day band; none of the three sources published a table like this on its own.
What the evidence says
Leah Schumacher and colleagues, publishing in the journal Obesity in 2019, tracked 375 members of the National Weight Control Registry (NWCR) who reported moderate-to-vigorous exercise on two or more days a week. NWCR members are a self-selected group who have each maintained a loss of at least 30 pounds for a year or more; the authors note in their limitations that entry is nonrandom and the group is demographically narrow, so the numbers below may not generalize past this specific population. Participants were split into temporally consistent exercisers, meaning more than half of a person's weekly exercise sessions fell inside one of four time bands (early morning, 4:00–8:59 a.m.; late morning, 9:00–11:59 a.m.; afternoon, noon–4:59 p.m.; evening, 5:00 p.m.–3:59 a.m.), and temporally inconsistent exercisers, whose sessions spread across bands. The consistent group, 68.0% of the sample, logged more active days a week (4.8 vs. 4.4, p = 0.007), more total weekly minutes (350 vs. 285, p = 0.03), and met the 150-minutes-a-week activity guideline more often (86.3% vs. 74.2%, p = 0.004). Early morning was the single largest band inside the consistent group, at 47.8%. The same paper also directly tested whether that specific time-of-day band related to activity levels, and found no significant difference between bands (days per week p = 0.06; weekly minutes p = 0.91; meeting the guideline p = 0.73). The comparison the study reports is between consistent and inconsistent timing overall, across all four bands.
Ben Edwards and colleagues published a smaller, lab-based study in Chronobiology International in 2025. Twelve recreationally active men completed three practice time trials, then six actual 10-km time trials, one at each of 06:00, 10:00, 14:00, 18:00, 22:00 and 02:00, spaced a week apart. A cosine model fit a significant 24-hour rhythm to their finish times (p < 0.05), a swing of about 5.0% across the day (d = 0.80), with the researchers reporting full statistical power for the result. The fitted curve's peak, called the acrophase, landed at 18:44, close to early evening; the authors report faster split times in the 18:00 and 22:00 trials than at 06:00, and describe performance as peaking in the evening. A parallel cosine fit on the runners' resting ear-canal temperature also showed a significant rhythm, consistent with the standard explanation that muscle and cardiovascular performance track the daily temperature cycle. The authors are explicit about the limits of their own sample: all 12 participants were men, and the paper states plainly that whether the same pattern holds in other populations is unclear.
Cleveland Clinic cardiologist Erik Van Iterson, PhD, addressed the same question for a general audience in a Cleveland Clinic article published in July 2024. The article itself states that afternoon or evening exercise is associated with peaks in body temperature and flexibility and appears better for aerobic performance and strength, a direction that matches the Edwards finding above. Separately, Van Iterson is reported as pointing out that exercising in the morning may be associated with lower blood pressure and better sleep. On the underlying question this page is built around, Van Iterson is quoted directly: “…for most of us who want to exercise for general health effects, the best time is the time that works best for your lifestyle and allows you to perform it regularly,” and, on picking one and holding it, “Regardless of why you exercise, pick the time that works best for you and stick with it.” The article's own narration concludes that consistency is the variable most strongly linked to good results, ahead of time of day, location or activity type.
How to apply this to your own schedule
Start by ruling out the hours that will not survive a normal week: a slot that depends on nobody else needing the shower, the car or the childcare at the same time tends to last past week one. Hold a single test slot for two to three weeks before switching to a different one. Schumacher et al.'s temporally consistent group in the numbers above was defined by holding to one time band across more than half of a week's sessions; the definition has room for the odd shifted run inside that band. If a run has to move from evening to a first attempt at morning, expect it to feel harder at the same easy pace for a week or two. The Cleveland Clinic article's note above, that body temperature and flexibility tend to peak later in the day, is one plausible explanation for that feeling, and it settles as the body adjusts.
Compare this week's chosen slot to your own last week, not to somebody else's ideal hour; in Runked the rank starts everyone at Bronze and only moves on your own runs. See how the running rank works.
Common mistakes
- Picking the hour a lab study calls fastest over the hour that actually fits a normal week. Edwards et al.'s 5% evening effect came from 12 men in one small trial, which the authors themselves call unclear whether it generalizes. Schumacher et al.'s adherence numbers came from a different measure entirely: how many people held to one time band, across a much larger sample.
- Switching time slots every few days. Schumacher et al. defined "temporally consistent" as more than half of a week's sessions falling in one band; switching daily keeps a runner inside the inconsistent group in that comparison, the group that logged fewer active days and lower total weekly minutes above.
- Assuming one hard first week at a new hour means it will not work. Van Iterson's own guidance above is to pick a slot and hold it regularly. None of the sources above measured outcomes from a single week.
What to do next
Picking a slot is easier once the rest of a training structure is in place. How to make running a habit covers the broader routine-building side of this question, including cues and habit stacking. How the running streak works covers what credits a day inside Runked's own streak count. And once a slot is chosen, how often should a beginner run is the next practical question.
Find out where you stand
Runked is free on the App Store. Connect your watch or track 3 runs, and meet your Runk.
Download on App StoreFrequently asked questions
Is running in the evening actually faster than running in the morning?
In one study, yes, with real caveats. Edwards and colleagues had 12 recreationally active men run six 10-km time trials at 06:00, 10:00, 14:00, 18:00, 22:00 and 02:00, one week apart each, published in Chronobiology International in 2025. A cosine model fit a significant 24-hour rhythm in finish time (p < 0.05), a swing of about 5.0% across the day (effect size d = 0.80), with the fitted curve's peak, its acrophase, landing at 18:44. The authors describe the sample as small and say it is unclear whether the pattern holds in other groups, such as women or less experienced runners.
What if the only time I have to run is early morning or late at night?
Leah Schumacher and colleagues tracked 375 members of the National Weight Control Registry, a self-selected group who had each kept off at least 30 pounds of weight for a year or more, publishing the results in the journal Obesity in 2019. Exercisers who reported training at the same time of day in more than half of their weekly sessions logged more active days per week (4.8 vs. 4.4, p = 0.007), more total weekly minutes (350 vs. 285, p = 0.03), and met the 150-minutes-a-week activity guideline more often (86.3% vs. 74.2%, p = 0.004) than exercisers whose timing varied week to week. Early morning, 4:00 to 8:59 a.m., was the single largest time band among the consistent group, at 47.8%. The same paper found no significant difference in activity levels between specific time bands (days per week p = 0.06; weekly minutes p = 0.91; meeting the guideline p = 0.73). The authors also flag that this self-selected, demographically narrow sample may not generalize to other groups.
Does running in the morning make a running habit easier to stick with?
Not by itself. Early morning, 4:00 to 8:59 a.m., was the single largest time band in Schumacher and colleagues' consistent-timing group, at 47.8% of that group. But the same paper directly tested whether that specific time-of-day band related to activity levels and found no significant difference between bands (days per week p = 0.06; weekly minutes p = 0.91; meeting the guideline p = 0.73). The adherence figures elsewhere on this page compare consistent timing overall against inconsistent timing, across a mix of four bands. They do not isolate morning as the cause.
Does Runked care what time of day I run?
No. Free-tier tracking, rank, streaks and the weekly effort-scored leagues, which reset every Monday at 00:00 UTC, work off which runs happened and how hard they were; none of them requires a specific clock hour. PRO's training plans schedule a run to a day inside a phase, such as Foundation or Race Prep, and leave the exact time up to the runner.