Does the time of day matter for red light therapy?

“Use it in the morning, when your mitochondria are most active.” You will find some version of that advice on plenty of red light therapy sites, sometimes wrapped in a story about ancient humans and sunrise. Where does this come from, and does it hold up? Out of thousands of studies, we found only six comparing night and day. They come from two laboratories and point in opposite directions.

Where “morning is best” comes from

The claim that morning is best rests on a single human study focussed on eyesight. In Shinhmar et al., 2021, researchers gave participants three minutes of 670 nm light to the eyes (HeroPanels Elite features 675 nm, a practical overlap with 670) and measured colour-contrast vision. Exposure between 8 and 9 in the morning improved colour discrimination by 12-17% for about a week. Exposure between noon and 1 p.m. did nothing. So for eyesight, it appears morning is superior.

A few caveats to keep in mind: This result came from 20 people in the morning arm and six in the midday arm, with different people in each arm rather than the same people tested twice. It was one laboratory, one tissue - the retina - and there has been no independent replication in the five years since. So this must be taken with a grain of salt.

A more broad cellular improvement was shown in fruit flies. The same group found that 670 nm light raised respiration and ATP by more than 40% at dawn but not eight hours later (Weinrich et al., 2019). Another study found that ATP rose by 20% at 08:00 and 15% at 11:00, with no effect at 05:00 or 16:00 (Shinhmar et al., 2022). Interestingly, the flies’ mitochondrial enzyme activity peaked in the afternoon - when the light did nothing. A loosely parallel (and highly speculative) finding comes from mice, where LED light raised muscle ATP and fatigue resistance most when given 6 hours before exercise rather than 3 hours or 5 minutes before (Ferraresi et al., 2015) - though human trials of the same idea have been mixed (Rossato et al., 2018; Abreu et al., 2020).

The authors’ proposed explanation is “spare capacity”: in the morning, energy stores are full and the mitochondria are idling, so light can push them harder; by afternoon they are already working flat out to replenish energy used during the day, leaving no capacity to utilise light. Spare capacity was never measured directly, and this is one study in flies - but it does mean the popular claim that “mitochondria are most active in the morning” gets the fly data backwards.

The studies that point the other way

The other timing comparisons come from a second laboratory, working in mice. Semyachkina-Glushkovskaya et al., 2021 found that 1267 nm light applied through the skull cleared amyloid faster and improved memory more when given at night or during sleep than during the day. The same group reported the same direction with 635 nm light during EEG-verified deep sleep (2023) and with 1050 nm light in aged mice, where waking treatment was ineffective and only treatment during sleep restored memory and metabolic markers (Terskov et al., 2025). Of course this also has its weaknesses; it's just one lab, three wavelengths, and completely based on animals not humans.

A 2022 review that tried to reconcile the two literatures found only two studies comparing day with night at the time. It described its proposed explanation - a mitochondrial mechanism while awake and a waste-clearance mechanism while asleep - as speculation.

That leaves the complete set of time-of-day comparisons in the history of photobiomodulation at one human eye study, two fly studies and three mouse brain studies, all from two laboratories, with conflicting results. For skin, muscle, wounds, hair, pain, immune function, metabolism and sleep, the number of studies comparing treatment times appears to be zero.

Are your mitochondria “most active in the morning”?

Not according to the strongest human evidence. van Moorsel et al., 2016 took muscle biopsies around the clock and found that oxidative capacity peaked at about 11 p.m. and troughed at about 1 p.m. - a swing of roughly 19%, and the reverse of the fly pattern. That finding is now contested: a 2025 paper titled “Missing the rhythm in skeletal muscle mitochondrial respiration” argues that the rhythm may not exist at all. It may also be absent in many of the people likely to be most at risk; Wefers et al., 2020 and Gemmink et al., 2023 found no daily rhythm in muscle mitochondrial capacity among older, overweight or glucose-intolerant adults.

Cytochrome c oxidase, the enzyme thought to absorb red light, does not appear to follow a clock either. In rat brain, it is more active during wakefulness or sleep deprivation than during sleep (Nikonova et al., 2005). It tracks arousal, not time of day. That would predict that light works better on an awake brain - the opposite of the mouse sleep results. The mechanism is still not settled in either direction.

Another result often gets attached to the morning story. Powner and Jeffery, 2024 found that 670 nm light on the upper back reduced the rise in blood glucose after a sugar drink by about 28% - a good result - and then wrote that “morning is the only time window when 670 nm exposure is effective”. The study did not test that claim: every session took place at the same time of day. The statement was carried over from the fly work, we believe it should be treated as speculative until evening/nighttime 670 nm light is shown not to be effective at reducing blood glucose after a sugary drink.

The evening question

The mirror-image concern is that red light in the evening will wreck your sleep. Here, the evidence is thin but points to caution. Ho Mien et al., 2014 exposed young men’s eyes to dim 631 nm red light in the evening. It did not suppress melatonin, but it delayed their body clocks by about an hour. This result is often misattributed to Figueiro and Rea, whose red-light work measured alertness and cortisol, not clock shifting. Red light at night is, in fact, used deliberately to keep shift workers alert.

In rodents, Stritzel et al., 2025 found that 634 and 658 nm light activated the master clock and suppressed melatonin at just a few microwatts per square centimetre - tens of thousands of times below a panel’s output - while 935 nm light did neither. Against that sits the only human evening study: Zhao et al., 2012: Twenty basketball players who lay under a red-light bed every evening for two weeks reported better sleep and higher melatonin than a control group. The control group received no light at all, making the results contradict the idea that red light at night may still disrupt sleep (though due to the nature of light being obvious to the senses, this lack of a blind control group also means placebo can't be ruled out). We examine red light therapy's interaction with circadian rhythm and melatonin in more detail in our post on sleep. Overall, the rodent and human results conflict and nobody has tested whether a real panel, used at a real distance, shifts the human body clock. Avoiding late-evening red light is therefore a precaution, not a response to an established harm, and the basketball player study suggests there may be no need to avoid it at night at all.

The obvious experiment nobody has run

Ironically, the best-established circadian effect in this area has never been connected to light therapy. Hoyle et al., 2017 found that burns sustained at night took about 40 days to reach 95% healing, compared with about 25 days for daytime burns, because the cells that close wounds run on a clock. Whether red light can alter that - or whether it works better on a wound at one time than another - has never been tested. It is arguably the most obvious untested hypothesis in the field. Maybe red light could close that enormous 15 day healing gap by simulating the wound as having occurred during the day?

What we tell people

For almost every use, the best time for red light therapy is the time you will use it consistently. Adherence matters more than any hypothetical timing effect, and most of the trials behind our dosing guide did not record the hour of treatment at all.

For the eyes specifically, morning is a reasonable, cheap and low-risk default. The one human timing comparison supports it, and the fly work offers a plausible mechanism. For later sessions, a sensible hedge is to avoid just before bed though many users (including me) find it still improves their sleep.

Do not believe anyone who says morning is best for skin, muscle, hair, pain, immunity or metabolism: no study has compared treatment times for any of them. Do not believe that red light therapy resets your body clock or treats jet lag: there is no primary literature behind either claim, though one small unblinded trial found it improved sleep in shift-work nurses with insomnia (Liao et al., 2025).

The bottom line is that time of day has scarce evidence as a significant variable in red light therapy, so choose the time which works best for your lifestyle.