It was 2:47 a.m. on Night Four when I stood in my kitchen, hands shaking from exhaustion, staring at my coffee maker and calculating whether I could survive the next workday without it. Sleep restriction therapy had limited me to a 5.5-hour sleep window—midnight to 5:30 a.m.—and I'd just spent the last two hours lying awake, watching the ceiling. The alarm would go off in less than three hours. Every part of me wanted to quit, to go back to my old pattern of crawling into bed at 9 p.m. and hoping for the best.
That moment, more than any other, is when I understood why most people never make it through CBT-I.
The Three Years Before
I'd been fighting insomnia since my early thirties. The usual story: stress at work triggered a few bad nights, which became a few bad weeks, which somehow became a permanent feature of my life. I tried everything that counts as "sleep hygiene." Blue light blockers. Magnesium supplements. A $600 mattress topper. Meditation apps. Chamomile tea that tasted like yard clippings.
Some nights I'd fall asleep fine but wake at 3 a.m., mind racing. Other nights I'd lie there for two hours before drifting off, then wake exhausted at 7 a.m. My average was maybe five hours of broken sleep, and I spent most of my waking life either exhausted or anxious about the next night.
When a friend mentioned CBT-I—cognitive behavioral therapy for insomnia—I looked into it. The research from Arthur Spielman and colleagues at CUNY was solid: 70-80% of people see significant improvement. But when I read about sleep restriction, the core component, my first thought was: this is completely backwards.
The Insane Part
Sleep restriction therapy asks you to do the opposite of what every instinct tells you. You're exhausted? You need more sleep? Too bad. You're going to get less.
Here's how it works: You track your sleep for a week to see how much you're actually sleeping (not just lying in bed). Let's say you're in bed nine hours but only sleeping five. Sleep restriction says: your new sleep window is five hours and fifteen minutes. That's it. No naps. No lying in bed outside that window. You're essentially consolidating all your sleep into one tight block and building up sleep pressure like a dam.
I started with a midnight-to-5:30 a.m. window. Before that first night, I felt genuinely afraid. What if I didn't sleep at all? What if I made things worse? I'd spent three years trying to get more sleep, and now I was voluntarily signing up for less.
Week One: The Hardest Seven Days
Night One, I was awake until 1:15 a.m. Got maybe four hours total. Night Two, I fell asleep by 12:30 a.m. but woke at 4 a.m. and couldn't get back. Night Three was better—asleep by 12:20 a.m., only woke once. Then came Night Four, the kitchen scene, where I genuinely considered quitting.
The protocol was clear about this part, and I held onto it like a lifeline: if you're awake more than 15-20 minutes, get out of bed. Don't lie there spiraling. So at 2:47 a.m., I got up. I sat in my dim living room with a book. Not a phone—an actual paper book. I was too tired to focus on the words, but I stayed there until I felt genuinely sleepy again. Back to bed at 3:40 a.m. Asleep by 3:55 a.m. Alarm at 5:30 a.m.
Two hours of sleep before a full workday.
I don't remember much of that Friday except drinking too much coffee and feeling like I was moving through wet concrete. But I made it to midnight without napping. And that night—Night Five—I was asleep within ten minutes.
The Moment It Clicked
By the end of Week One, my sleep efficiency—time asleep divided by time in bed—had jumped from 58% to 79%. I was still only getting 5.5 hours, but it was consolidated. No more lying awake for hours. No more 3 a.m. anxiety spirals. When I got into bed at midnight, my body knew what to do.
Week Two, the window expanded to six hours. Midnight to 6 a.m. The rule was simple: once you're sleeping 85% or more of your time in bed for five consecutive nights, you add 15 minutes. It felt like earning something back.
That first night with the extra 30 minutes, I slept straight through. Six solid hours. I woke up at 6 a.m. and actually felt rested—a sensation I hadn't experienced in so long I'd almost forgotten what it was.
By Week Three, I was up to 6.5 hours. By Week Five, seven hours with 88% efficiency. My body had relearned what bed was for. Not worrying. Not checking the clock. Sleeping.
What Changed
The mechanics made sense in hindsight. For years, I'd been spending nine hours in bed trying to catch five hours of sleep. Every night, my brain learned: bed equals wakefulness, anxiety, frustration. Sleep restriction broke that association by making bed a place where I was simply too tired not to sleep.
There were other pieces that helped. The stimulus control rules: no phones in bed, ever. If I couldn't sleep, I left the room—no negotiation. And the cognitive work: noticing the catastrophic thoughts ("I'll never sleep again") and letting them pass instead of engaging.
But sleep restriction was the engine. Everything else was fine-tuning.
I followed the full protocol from Rest →, which laid out the entire six-week process with specific sleep window calculations, expansion schedules, and what to do when you hit plateaus (I hit two). It wasn't easy. There were nights I wanted to quit. But having the structure—knowing exactly when to expand the window, what efficiency threshold to aim for—made it doable.
Six Months Later
I'm in bed from 11:15 p.m. to 6:45 a.m. most nights. I sleep about seven hours with 90% efficiency. Some nights are still harder than others—stress, travel, the occasional bout of noise from the neighbors—but I don't spiral anymore. I know what to do: tighten the window for a few nights, rebuild the pressure, let my body remember.
The hardest part wasn't the exhaustion of Week One. It was trusting that something so counterintuitive could actually work. But that's the thing about sleep restriction: it only sounds insane until you try it.
— Simon