Experiment guides → When sleep goes wrong → Tired but wired
Why can't you fall asleep even when you're tired?
Quick answer: because tiredness and sleepiness aren't the same thing. Tiredness is a lack of energy; sleepiness is the brain's readiness to switch off — and you can have plenty of the first while something holds the second down. Sleep researchers call the state hyperarousal: sleep pressure is high, but the alerting system is higher. The common things holding it up are boringly physical: caffeine that's still active eight hours after your last cup, bright screens pushing the body clock later, a wake time that drifts so the clock never settles, or a bed that has learned to mean "lie here and think". Each has a 14-night experiment. Run the one that fits your evening first.
What the research says
Hyperarousal: the model that explains tired-but-wired. Riemann's review sets out the evidence that insomnia involves elevated arousal — cognitive, emotional and physiological — around bedtime and across the 24 hours: faster brain-wave activity at sleep onset, raised heart rate and cortisol, a mind that won't idle. Sleep pressure builds normally; it just can't win. Treatments that lower arousal (stimulus control, CBT-I, relaxation) work on exactly this (Riemann et al., 2010).
Suspect one: caffeine you'd forgotten about. A 2023 meta-analysis puts caffeine's average cost at 45 minutes of sleep and 7% of sleep efficiency, and estimates a standard cup needs about 8.8 hours before bed to leave sleep untouched. A 3 PM coffee is still working at 11 PM in many people — and in a controlled trial, caffeine six hours before bed cut measured sleep by over an hour without the sleepers noticing (Gardiner et al., 2023).
Suspect two: evening light, holding the clock open. Four hours of a bright screen before bed suppressed melatonin by about 55% and delayed the body clock by about 1.5 hours in a crossover study; the same sleep pressure, a clock that says it's not night yet. Ordinary scrolling is a smaller dose, but it's in the same direction, and it adds content-driven arousal on top (Chang et al., 2015).
Suspect three: stress reactivity — some brains are wired for it. Kalmbach's review of sleep reactivity describes a trait: people whose sleep is easily disrupted by stress, who lie awake after a difficult day even when exhausted. It's partly heritable and it predicts who develops insomnia. For them, the experiments that lower arousal — a to-do list, breathing, stimulus control — matter more than the ones that move the clock (Kalmbach et al., 2018).
What this evidence doesn't say
This page covers the common lifestyle causes in otherwise healthy people. Trouble falling asleep that has lasted more than three months, several nights a week, with daytime consequences, is chronic insomnia — a condition with a first-line treatment (CBT-I) that beats any single experiment. Restless legs, anxiety disorders, thyroid problems and some medications also present as tired-but-wired and need a clinician.
Pick your suspect, then run one experiment
- Five nights of baseline, with an evening audit. Each morning: sleep quality, restedness, time-to-sleep. Each evening: last caffeine, last screen, and a 1–5 score for how wired you felt at lights-out — mind racing, body tense, or both.
- If you had caffeine after lunch: the cutoff first. It's the biggest documented lever and the one people underestimate most. Run the caffeine cutoff — last cup by early afternoon, same total dose, 14 days.
- If the phone is in your hands at lights-out: screens. Run the screen-free hour or, simpler, phone out of the bedroom. Both remove light and content in one move.
- If your mind races: the to-do list, then breathing. The five-minute to-do list has a polysomnography result behind it and costs nothing; slow breathing is the next block if the mind is calm but the body isn't.
- If you've been lying awake in bed for weeks: stimulus control. Bed only for sleep, up after 20 minutes awake, fixed wake time — the full protocol. It's the treatment-grade experiment, and the one to run if the others have failed.
What to expect
Caffeine and light produce fast results when they're the cause: within a few nights, sleepiness arrives at bedtime instead of after midnight. Arousal-driven wiredness takes longer, because the mind has to learn that bed is safe to switch off in; the stimulus-control block often gets worse for a week before it gets better. If none of the four changes anything in six to eight weeks of testing, that's the point to seek CBT-I rather than a fifth experiment.
Who should skip this
See a doctor rather than experimenting if you have creeping or restless sensations in your legs at night, if the wiredness comes with a racing heart, weight loss or heat intolerance, if you've started or changed a medication, or if anxiety is present by day as well. Chronic insomnia deserves treatment, not a longer list of things to try.
Frequently asked questions
Why am I exhausted but can't sleep?
Exhaustion is low energy; sleep needs low arousal. Caffeine, evening light, stress and a conditioned bed all keep arousal up while energy is down — the tired-but-wired state. The fix is lowering arousal or removing what's raising it, not adding tiredness.
How long does caffeine keep you awake?
Longer than it feels: a meta-analysis estimated a standard cup should be finished about 8.8 hours before bed to avoid cutting sleep, and a controlled trial found 400 mg six hours before bed still cost over an hour of measured sleep — mostly unnoticed by the sleepers. Metabolism varies several-fold between people, which is why a cutoff test beats a rule.
What is sleep hyperarousal?
The leading model of insomnia: a state of elevated alertness — faster brain activity, raised heart rate and cortisol, an active mind — that persists at bedtime and across the day. Sleep pressure is normal; the brake on sleep is stronger. It's why sleep-inducing tricks often fail and arousal-lowering treatments (stimulus control, CBT-I) succeed.
Related experiments
Sources
- Riemann D, Spiegelhalder K, Feige B, et al. The hyperarousal model of insomnia: a review of the concept and its evidence. Sleep Med Rev. 2010;14(1):19–31. PubMed 19481481
- Gardiner C, Weakley J, Burke LM, et al. The effect of caffeine on subsequent sleep: a systematic review and meta-analysis. Sleep Med Rev. 2023;69:101764. PubMed 36870101
- Chang AM, Aeschbach D, Duffy JF, Czeisler CA. Evening use of light-emitting eReaders negatively affects sleep, circadian timing, and next-morning alertness. Proc Natl Acad Sci U S A. 2015;112(4):1232–1237. PubMed 25535358
- Kalmbach DA, Cuamatzi-Castelan AS, Tonnu CV, et al. Hyperarousal and sleep reactivity in insomnia: current insights. Nat Sci Sleep. 2018;10:193–201. PubMed 30046255
Track this experiment in N of 1
A few baseline nights, then the change — and your own mornings deliver the result. No streaks, no guilt. Your first experiment is free.
Downloadon the App StoreThis page is general information about published research, not medical advice. If you have a sleep disorder, are pregnant, or take medication, talk to your doctor before changing your routine.