Getting the phone out of the bedroom, and why blue light is the least of it

What the blue light study really showed

The reference behind nearly every phones-and-sleep article is Chang, Aeschbach, Duffy and Czeisler, published in PNAS in 2015, out of Brigham and Women's Hospital and Harvard Medical School. It is a good study and the findings are striking.

Reading on a light-emitting eReader in the evening, against a printed book, suppressed evening melatonin by about 55%, pushed dim-light melatonin onset more than 1.5 hours later, lengthened the time taken to fall asleep, reduced and delayed REM sleep, and left people measurably less alert the next morning.

Now the part that gets left out

Participants read for four hours before bed, in a controlled laboratory, on a device held at reading distance in an otherwise managed light environment. That is a deliberately strong exposure, designed to produce a signal big enough to measure.

It is not the same experiment as checking your phone for two minutes at 1am. Nothing in the paper identifies a threshold below which the effect disappears, and the popular one-hour rule is a practical compromise rather than the study's result. I went through this in more detail when tracing the 10-3-2-1-0 rule, whose screen step rests on the same paper.

So: the light effect is real, well measured, and almost certainly overstated as a reason to remove your phone from the room. If light were the whole story, an e-ink reader or a dim screen would solve it, and in my experience it does not.

The reason that holds up

The better argument is one I only found by tracing something else, and it turns out to be fifty years old.

Stimulus control therapy, formalised by the American psychologist Richard Bootzin in 1972, is the only nonpharmacological insomnia treatment holding the highest standard-of-practice rating from the American Academy of Sleep Medicine. Its instructions include: go to bed only when sleepy, get out of bed if you are not asleep, get up at the same time each morning, and use the bedroom for sleep and sex only.

That last instruction is the phone rule, written before phones existed.

The mechanism is conditioning. Your brain associates places with states. A bed used only for sleep becomes a cue for sleepiness. A bed used for scrolling, working, worrying and watching becomes a cue for being alert, and the association is built by repetition rather than by intention. It does not care that you meant to sleep.

A phone on the nightstand guarantees the wrong repetition. Not because you plan to use it, but because at 2am, awake and bored, you will, and each time you do you are teaching your brain something about what beds are for. The full protocol is here.

My own numbers

Across the 28 nights of February 2026 I logged whether I had used a screen in the hour before bed, alongside an estimate the next morning of how long I took to fall asleep.

Seven minutes of median difference, on my own nights. I find that persuasive and I am not going to oversell it.

It is one person. Latency was estimated the next morning rather than measured, and self-reported latency skews long. Twenty-eight nights split by this factor leaves few observations either side. And most importantly, screen nights were not otherwise identical nights: they correlated with higher stress and with later caffeine, both of which had larger gaps in my data than screens did. Stress alone was 24 minutes against 9.

So the honest version is that screen use marks a kind of night rather than proving a mechanism. It still moved me, because the direction agrees with the conditioning argument and the cost of acting on it was one cheap alarm clock. Full distribution and caveats here.

How to actually do it

• Charge it in another room. This is the whole intervention. Everything else is a way of avoiding it.
• Buy an alarm clock. Five pounds, and it removes the only genuinely good reason to keep the phone by the bed.
• Sort out the emergency case first. If someone might need to reach you at night, set up a bypass that rings through and put the phone somewhere audible but not reachable from bed. Solve it once rather than using it as a reason not to start.
• Move the charger, not just the phone. If the cable is by the bed, the phone comes back within a week.
• Do not negotiate face-down. Face down on the nightstand is the version that fails, because reachable is the property that matters.

The objections

"I use it as my alarm." The most common one and the most easily solved. An alarm clock costs almost nothing and does one job.

"I need it for emergencies." Legitimate, and worth actually solving rather than using as a reason. Most phones let specific contacts ring through. Audible from the bedroom, not reachable from the bed.

"I read on it." Then the conditioning argument applies to reading too, which is genuinely inconvenient. The protocol's answer is to read somewhere else and go to bed when sleepy. Mine is that a paper book in bed is a compromise I make, and I am telling you it is a compromise rather than pretending it is compliant.

"I sleep fine with my phone there." Then this is not your problem and you should ignore this page. Stimulus control is a treatment for a specific failure mode. If your bed is already a reliable sleep cue, protecting it further has little to add.

Questions people actually ask

References

• Chang AM, Aeschbach D, Duffy JF, Czeisler CA. Evening use of light-emitting eReaders negatively affects sleep, circadian timing, and next-morning alertness. PNAS. 2015;112(4):1232–1237. Source for the 55% melatonin suppression, the 1.5 hour delay, and the four-hour exposure.
• Bootzin RR. Stimulus control treatment for insomnia. 1972. Source for the bedroom-for-sleep-only instruction and the conditioning rationale.
• American Academy of Sleep Medicine practice standards for behavioural treatment of insomnia. Source for the standard-of-practice rating.
• Self-tracking diary, 1 to 28 February 2026. Source for the 16 against 9 minute comparison. Method and limitations on the nap calculator.

Related reading

• The 20-minute rule, the rest of the protocol this instruction belongs to.
• Five things I changed to sleep better, and how good the evidence is for each.
• Sleep regularity beats sleep duration, the change with the strongest evidence behind it.
• The 10-3-2-1-0 rule, whose one-hour screen step rests on the same PNAS paper.