Why You Can’t Sleep Even Though You’re Exhausted

It’s 2:47 AM. You know the exact time because you’ve checked your phone again, which means you’ve blasted your retinas with blue light again, which means you’ve reset the clock on falling asleep again, which you know, which makes it worse. Your body is lead. Your eyelids could be made of concrete. Every muscle has filed its resignation and gone home for the night. But your brain—your brain is running a TED Talk nobody asked for, and the audience is you, and there is no exit.

The topic keeps shifting. First it was tomorrow’s meeting. Then it was that thing you said at dinner three days ago that was probably fine but might not have been. Then it was finances. Then something existential and formless that you can’t quite name but that sits on your chest like a cat made of dread. Now it’s the fact that you’re not sleeping, and you have to be up in four hours, and if you fall asleep right now you’ll get—you’re doing the math again. You always do the math. The math never helps.

You flip the pillow. The cool side lasts about eight seconds.

Here’s what no one told you: this isn’t a failure of willpower. You can’t force yourself to sleep any more than you can force yourself to digest. Sleep is not something you do. It’s something that happens when certain conditions are met, and right now, your biology is actively preventing those conditions from being met. There’s a reason for it, it’s well understood, and the fact that you’re exhausted is not a contradiction. It’s the core of the problem.

What’s Actually Happening in Your Body

The intuitive model of insomnia is wrong. Most people assume they can’t sleep because they aren’t tired enough—that insomnia is an insufficiency of sleepiness. This is backwards. The dominant model in sleep science, the hyperarousal model proposed by Bonnet and Arand, establishes that insomnia is not a deficit of sleepiness. It’s an excess of arousal. You are tired. You are also wired. Both are true simultaneously, and the wired wins.

Your body is running on systems that evolved to keep you alive in environments where falling asleep at the wrong moment meant becoming something else’s dinner. These systems are spectacularly good at their job. The problem is they don’t know the difference between a lion in the grass and a quarterly review on Monday.

Your Stress System Won’t Stand Down

The hypothalamic-pituitary-adrenal axis—your HPA axis—is your body’s central stress command. In a well-regulated system, cortisol follows a clean diurnal rhythm: it peaks in the early morning to get you moving, declines through the day, and reaches its lowest point around midnight. This decline is not optional for sleep. It’s prerequisite.

In people with chronic insomnia, research shows the HPA axis stays overactive. Cortisol remains elevated in the evening and overnight when it should be at its nadir. This isn’t because you’re choosing to be stressed. The axis has become dysregulated—stuck in a mode where it overproduces cortisol even in the absence of acute threat. Your body is running its emergency generator when the power is fine.

Simultaneously, your sympathetic nervous system—the fight-or-flight branch—maintains elevated tone. Heart rate stays slightly higher than it should. Core body temperature doesn’t drop the way it needs to. (Falling asleep requires a core temperature decline of roughly 1-2 degrees Fahrenheit, and sympathetic activation opposes this.) Muscle tension persists at a low level you don’t consciously notice but your body absolutely does. You are, in a measurable physiological sense, too activated to sleep despite being too exhausted to function. The gas pedal and the brake are both pressed to the floor.

The Noise That Keeps You Awake

That TED Talk your brain is running at 3 AM? Neuroscience has a name for the network responsible: the default mode network (DMN). This is the constellation of brain regions that activates when you’re not focused on external tasks—it handles self-reflection, future planning, autobiographical memory, and the uniquely human talent for replaying embarrassing moments from 2014 in high definition.

In a healthy sleep transition, DMN activity quiets as you drift off. The internal monologue fades. Thoughts become less linear, more fragmented, more dreamlike. But in insomnia, particularly the kind driven by anxiety and rumination, the DMN doesn’t quiet. It persists. And the content it generates—worry, planning, self-criticism, ambient dread—is precisely the kind of cognitive activity that sustains arousal. You’re not choosing to ruminate. Your default mode network is generating thoughts automatically, and those thoughts are keeping the arousal system engaged, which keeps the DMN active, which generates more thoughts. It’s a feedback loop, and lying in the dark with nothing to distract you makes it louder, not quieter.

The Things You Did Today That Are Still Affecting You Tonight

Caffeine has a half-life of approximately 5-6 hours in most adults. This means that your 2 PM coffee—the one that felt necessary, the one that got you through the afternoon—still has 50% of its caffeine active in your system at 8 PM. By midnight, roughly 25% remains. This is caffeine’s quarter-life, and it rarely appears in the conversation about sleep, but it should. That 200mg afternoon latte is still delivering 50mg of stimulant at midnight. You wouldn’t take a caffeine pill at midnight. But effectively, you did.

Caffeine blocks adenosine receptors. Adenosine is the molecule that builds up during waking hours and creates sleep pressure—the biological signal that says “it’s time.” When caffeine occupies those receptors, the signal doesn’t arrive. Your brain has spent the day accumulating adenosine, your body is exhausted from using it, but your brain can’t hear the message telling it to sleep.

Alcohol is the other half of this trap, and it’s more insidious because it feels like it works. A glass of wine does help you fall asleep faster. The sedative effect is real. But research by Ebrahim and colleagues (2013) confirmed what sleep scientists have known for decades: alcohol fragments the second half of the night, suppresses REM sleep, increases nocturnal awakenings, and worsens sleep quality overall. The nightcap is a loan shark—it gives you 20 minutes of faster sleep onset and takes back two hours of restorative architecture. People who drink regularly to sleep often don’t realize that the alcohol is causing the fragmented sleep that makes them feel they need alcohol to sleep. The cycle is invisible from the inside.

Blue light deserves a mention but not a starring role. Yes, short-wavelength light from screens suppresses melatonin production by signaling to your suprachiasmatic nucleus that it’s still daytime. Yes, this delays sleep onset. But blue light is a piece of the puzzle, not the puzzle. People slept poorly long before smartphones. If you fix your light exposure but do nothing about cortisol dysregulation, rumination, and caffeine timing, you will still lie awake. The screen is the easy scapegoat. The nervous system is the harder truth.

When Not Sleeping Becomes the Problem

Here’s where insomnia gets recursive. After enough nights of lying awake, your brain begins to associate the bed itself with wakefulness. This is classical conditioning—the same mechanism Pavlov demonstrated with dogs and bells, except the bell is your pillow and the response is hyperarousal.

Sleep researchers call it conditioned arousal, and it’s the reason insomnia becomes self-sustaining even after the original trigger resolves. The work stress passes. The deadline ends. But now you have a new problem: the anxiety about not sleeping has become the thing preventing sleep. You get into bed and your heart rate ticks up. Not because anything is wrong. Because your brain has learned that this is the place where you don’t sleep, and it’s preparing accordingly. The bed has become a cue for vigilance rather than rest.

This is also why “trying harder” to sleep is counterproductive. Effort is arousal. Monitoring your own state—“Am I falling asleep yet? Why am I not falling asleep yet?”—is arousal. The paradox is that sleep requires a surrender of the very intentional control that everything else in your life rewards. And for people whose coping strategy for most problems is to try harder, this is maddening.

What Actually Helps

Not everything recommended for sleep is equally supported. Some interventions have decades of clinical evidence. Some have promising preliminary data. Some have centuries of traditional use and not much else. And some are actively harmful despite being popular. Here’s an honest sorting.

Strong Evidence

CBT-I (Cognitive Behavioral Therapy for Insomnia) is the single most effective long-term treatment for chronic insomnia, and it’s not close. A 2015 meta-analysis by Trauer et al. published in the Annals of Internal Medicine found that CBT-I produced durable improvements in sleep onset latency, wake-after-sleep-onset, total sleep time, and sleep efficiency—and these improvements persisted after treatment ended. Sleeping pills work while you take them. CBT-I teaches your brain to sleep again.

The core components: sleep restriction (counterintuitively, spending less time in bed to consolidate sleep drive), stimulus control (bed is for sleep and sex only—no reading, no scrolling, no lying awake), cognitive restructuring (addressing the catastrophic thinking about sleep that fuels conditioned arousal), and relaxation training. It typically runs 6-8 sessions. It is more effective than medication by every long-term measure, and it’s available through therapists, online programs, and even apps. If you take one thing from this article, make it this: CBT-I exists and it works.

Consistent wake time matters more than consistent bedtime, and this surprises people. Your circadian clock is anchored more strongly by when you wake up than when you go to sleep. Setting a fixed wake time—same time every day, weekends included—is the single most powerful circadian intervention available. It’s free, it requires only an alarm, and it works by stabilizing your entire cortisol-melatonin rhythm. Sleeping in on weekends feels like recovery. It’s actually circadian jet lag.

Temperature manipulation exploits the core temperature drop that sleep initiation requires. A 2019 meta-analysis by Haghayegh et al. found that a warm bath or shower taken 1-2 hours before bed significantly improved sleep onset latency and sleep quality. The mechanism is paradoxical: warming the skin surface causes vasodilation, which accelerates heat loss from the core, which drops core temperature faster than it would decline naturally. The warm bath doesn’t warm you up for sleep. It helps you cool down. Keep the bedroom cool—65-68 degrees Fahrenheit is the range most sleep research supports. Your body can’t achieve the thermal drop it needs in a warm room.

Specific sleep hygiene—and I mean specific, not the generic list you’ve already read. The version that actually matters: no caffeine after noon (or earlier if you’re a slow metabolizer—roughly 50% of the population carries CYP1A2 variants that slow caffeine metabolism). Get bright light exposure within 30 minutes of waking—this sets the cortisol peak early and programs the melatonin onset 14-16 hours later. If you’re awake in bed for more than 20 minutes, get up. Go to another room. Do something quiet and boring in dim light until you feel sleepy again, then return to bed. This is stimulus control, and it breaks the conditioned association between bed and wakefulness.

Good Evidence

Magnesium glycinate addresses a nutritional gap that’s more common than most people realize. An estimated 50% of North Americans don’t get adequate magnesium from diet. Magnesium activates the parasympathetic nervous system and binds to GABA receptors—the same receptor system that benzodiazepines target, with a vastly different risk profile. A 2012 randomized controlled trial by Abbasi et al. found that magnesium supplementation significantly improved insomnia severity, sleep time, and sleep efficiency. Dose: 200-400mg of elemental magnesium from glycinate form, taken 30-60 minutes before bed. The form matters—magnesium oxide, the cheapest version, has roughly 4% bioavailability. You want glycinate or citrate.

Passionflower (Passiflora incarnata) has a result that deserves more attention than it gets. In a 2011 double-blind study by Ngan and Conduit, passionflower tea consumed before bed for one week produced significant improvements in subjective sleep quality as measured by sleep diary and polysomnography. The mechanism: passionflower’s flavonoids act as positive allosteric modulators at GABA-A receptors, amplifying your brain’s natural GABA activity rather than overriding it. An earlier trial by Akhondzadeh matched passionflower against oxazepam—a prescription benzodiazepine—and found no significant difference in therapeutic effect, with fewer side effects in the passionflower group. A flower. Matching a benzo. In a randomized controlled trial.

L-theanine is the amino acid found primarily in tea plants, and it does something most calming compounds don’t: it relaxes without sedating. A 2019 randomized controlled trial by Hidese et al. found that 200mg of L-theanine daily reduced stress-related symptoms and improved sleep quality. The mechanism is alpha brain wave promotion—the pattern associated with relaxed wakefulness, the state during a good meditation session. L-theanine doesn’t knock you out. It quiets the conditions that keep you up. For the subset of people whose insomnia is primarily a “can’t stop thinking” problem, this distinction matters.

Ashwagandha (Withania somnifera) works on a different timeline and a different target. A 2019 randomized double-blind trial by Langade et al. found that 600mg of ashwagandha root extract daily for 10 weeks significantly improved sleep quality, sleep onset latency, and overall well-being compared to placebo. The mechanism is HPA axis modulation—ashwagandha is classified as an adaptogen, meaning it helps recalibrate the cortisol response rather than simply suppressing it. For people whose insomnia is rooted in chronic stress and cortisol dysregulation, ashwagandha addresses the upstream problem rather than masking the downstream symptom. The effects build over weeks, not hours. This is a recalibration, not a sedative.

Promising

Psilocybin microdosing occupies an interesting position in the sleep conversation because of what it doesn’t do. Users don’t report it as a sedative. They don’t describe feeling sleepy. What they consistently describe is that the mental noise—the rumination, the recursive worry, the 3 AM TED Talk—gets quieter. And when the noise gets quieter, sleep comes.

The mechanism has a plausible neurobiological basis. Carhart-Harris et al. (2012) demonstrated using functional MRI that psilocybin decreases activity in the default mode network—the same network responsible for the self-referential thinking, rumination, and worry that sustains insomnia-related arousal. When the DMN quiets, the cognitive interference that prevents the sleep transition diminishes. You don’t force sleep. You remove the obstacle.

Polito and Stevenson (2019) tracked microdosers over six weeks and found decreased mind-wandering and decreased depressive symptoms alongside increased absorption—the capacity to be present with what’s in front of you rather than pulled into recursive thought loops. For insomnia driven by rumination, these findings are directly relevant. Rumination is the engine. Quiet the engine, and the body does what it was trying to do all along.

The honest limitation: the large-scale, double-blind, placebo-controlled trials that would move microdosing from “promising” to “established” for sleep specifically haven’t been completed. The existing research is suggestive, mechanistically coherent, and consistent with user reports—but it’s not yet definitive. Anyone claiming certainty in either direction isn’t reading carefully. A comprehensive guide to microdosing protocols is a better starting point than anecdotes.

Traditional

Valerian root (Valeriana officinalis) has been prescribed for insomnia since the second century. The clinical evidence is genuinely mixed—meta-analyses show improvements in subjective sleep quality but inconsistent results on objective measures. The mechanism involves GABA modulation, the pharmacology is real, and centuries of use suggest something is happening, even if the clinical trial data is messier than we’d like. At $10-15 per month with minimal side effects, a two-week trial is reasonable. Chamomile and lavender occupy similar territory: long traditional use, plausible mechanisms (apigenin in chamomile binds GABA-A receptors; linalool in lavender has anxiolytic effects), and evidence that’s encouraging without being strong.

What’s Overhyped or Harmful

Melatonin overuse is the biggest misunderstanding in sleep supplementation. Melatonin is a timing signal, not a sedative. Your pineal gland produces roughly 0.1-0.3mg in the evening to signal “it’s nighttime” to your circadian system. Most supplements contain 3-10mg—ten to one hundred times the physiological dose. At these levels, you’re not supplementing a signal. You’re flooding a system. High-dose melatonin can cause next-day grogginess, vivid nightmares, and—paradoxically—worsened sleep quality over time as receptor sensitivity downregulates. For jet lag and shift work, short-term low-dose melatonin (0.3-0.5mg) makes physiological sense. For chronic insomnia, it’s treating the wrong problem. Your clock isn’t broken. Your arousal system is.

Sleep tracking obsession—what researchers have termed orthosomnia—is the ironic condition where anxiety about sleep data worsens sleep. Your watch says you got 47 minutes of deep sleep and now you’re worried that’s not enough, and that worry activates the arousal system, which will ensure you get even less tonight. The devices measure movement and heart rate, not actual sleep stages—their accuracy for staging is limited at best. If your tracker is making you more anxious about sleep, take it off. The data isn’t worth the cost.

Diphenhydramine (Benadryl, ZzzQuil) is the most common over-the-counter sleep aid in North America, and it deserves more caution than it gets. It works by blocking histamine receptors, which causes drowsiness. But it also has significant anticholinergic properties, and long-term anticholinergic use has been associated with increased risk of dementia in a 2015 study published in JAMA Internal Medicine. The study found a dose-response relationship—more cumulative use correlated with higher risk. Tolerance develops within days, meaning you need increasing doses for the same effect, and next-day cognitive impairment is well-documented. As an occasional tool, it’s fine. As a nightly habit, the risk profile is not what most people assume.

What We’d Actually Tell a Friend

If someone I cared about told me they hadn’t slept well in weeks and they were starting to feel like something was wrong with them, here’s what I’d say.

First: nothing is wrong with you. Your nervous system is doing exactly what a nervous system does when it’s been running in threat-detection mode for too long. Insomnia is a signal, not a malfunction. The signal is that your arousal system needs to be recalibrated, and that’s a solvable problem.

Second: start with the architecture, not the supplements. Fix your wake time. Make it the same every day for two weeks and don’t negotiate with yourself about weekends. Stop caffeine by noon—genuinely by noon, not “early afternoon.” Take a warm shower 90 minutes before bed. Keep the room cool and dark. If you’re in bed awake for 20 minutes, get up. These are not novel recommendations. They are relentlessly validated ones, and most people skip them because they feel too simple to work. They work.

Third: look into CBT-I. It’s the most effective treatment for chronic insomnia that exists. More effective than any pill by every long-term measure. If you can’t access a therapist, apps and online programs deliver most of the benefit.

Fourth: if you want supplement support, the evidence hierarchy matters. Magnesium glycinate is the easiest starting point—cheap, safe, addresses a common deficiency, and the effect on sleep is often noticeable within a week. Passionflower and L-theanine are the next tier, particularly if your insomnia is anxiety-driven. Ashwagandha if you’re willing to commit to the timeline for cortisol recalibration. A deeper look at these compounds and how they interact is worth the read before you start stacking.

Fifth: if the rumination is the main event—if what keeps you awake is not physical discomfort but the inability to stop thinking—that’s worth paying attention to as its own category of problem. Meditation helps. So does journaling before bed (externalize the thoughts so your brain stops holding them in working memory). And the emerging research on psilocybin microdosing and default mode network activity suggests a mechanism that specifically addresses cognitive noise, not through sedation but through quieting the source.

Sixth: if insomnia has been persistent, worsening, or accompanied by daytime impairment that’s affecting your work or relationships, see a sleep specialist. Not a general practitioner with a prescription pad—a specialist. Sleep apnea, restless leg syndrome, circadian rhythm disorders, and depression all present as “I can’t sleep,” and they require different interventions. Getting the right diagnosis is worth more than every supplement on this list combined.

The thing about insomnia is that it makes you doubt yourself. You start to wonder if you’ve forgotten how to do something that infants manage without instruction. You haven’t forgotten. The mechanism is intact. It’s being overridden by a system that evolved to protect you and doesn’t know the threats aren’t physical anymore. The biology that’s keeping you awake is the same biology that can be redirected to let you sleep.

It’s not about trying harder. It’s about creating the conditions where trying isn’t necessary.