Sleep Diary

The diary

Press Start a new 14-day diary to populate today and the next 13 days. Fill it in each morning — five minutes is enough. Your data is auto-saved to this browser as you type. When the two weeks are complete, use Print / Save as PDF or Export CSV — for a clinician visit, or just to keep as your own record of how you’ve been sleeping.

What a sleep diary tracks (and what it deliberately doesn't)

The fields above are the standard sleep-diary set used by the American Academy of Sleep Medicine, the consensus sleep-diary recommendation from the Pittsburgh and Penn sleep groups, and the diary template most cognitive-behavioral-therapy-for-insomnia (CBT-I) protocols use. They cover the four signals a clinician — or you, looking at your own data — will look at first:

• Sleep timing. Bedtime and wake-up time describe when your sleep window sits relative to the day. A consistent window across all 14 days is one signal; a chaotic or weekend-shifted window is a different signal.
• Sleep continuity. Time to fall asleep (sleep onset latency) and number of wake-ups during the night describe how fragmented the sleep window is. The two together separate "I sleep through but not long enough" from "I'm in bed for nine hours but only asleep for six."
• Subjective quality. The 1–5 quality rating captures how rested you actually feel. It frequently doesn't match the duration — people report "great" 6-hour nights and "terrible" 9-hour nights regularly. Both numbers tell you something different.
• Behavioral inputs. Caffeine, alcohol, and exercise are the three lifestyle factors with the largest effects on sleep architecture for most people. Note timing and amount roughly — exact precision isn't necessary.

What the diary deliberately doesn't track: anything that requires instrumentation. There's no field for heart rate variability, blood-oxygen saturation, REM-stage estimation, "sleep score" out of 100, or any of the metrics consumer wearables produce. Those numbers can be useful in their own right, but they're a separate evidence stream from the diary, and conflating them tends to dilute both. The diary is your subjective record. The wearable, if you use one, is the device's record. They're most useful side by side, not blended into one narrative.

The diary also doesn't ask for medication doses, mood ratings, or a detailed dietary log. Those belong in their own logs if a clinician asks for them, or if you find them personally useful. Keeping the diary focused on the four signals above means it stays a five-minute morning task instead of a project.

Why two weeks — not one, and not four

The 14-day window is a deliberate compromise. One week of data is too short to distinguish weekday-versus-weekend variation from a true pattern, and any single anomaly (a sick night, a travel night, an exam-week night) skews the average heavily. Four weeks is enough to surface real seasonal or behavioral changes, but adherence drops sharply past two weeks for most people — diaries that aren't filled in every morning become diaries that aren't reliable.

Two weeks gives roughly ten "typical" nights after accounting for one-off disruptions, captures both weeknights and weekend nights, and is short enough that most people actually complete it. The American Academy of Sleep Medicine clinical guideline for chronic insomnia specifically recommends a two-week sleep diary as part of the initial evaluation, and CBT-I protocols use the same two-week window for baseline establishment before behavioral changes begin.

If the two weeks reveal something concerning enough that a clinician wants more data, they may ask for a second two-week stretch, or pair the diary with actigraphy (wrist-worn movement tracking) or a home sleep apnea test. Those are deliberate next steps under clinical supervision, not extensions of the self-tracking exercise. If you’re using the diary for your own purposes, two weeks is also enough to establish a personal baseline before trying out a behavioral change (cutting caffeine after noon, holding wake time constant on weekends, dimming lights in the last hour) and seeing whether it moves the numbers.

How to fill it in

Fill it in in the morning, not at night. Morning entries are more accurate because the night's events are still fresh and you're not estimating in advance. Set a phone reminder for fifteen minutes after your usual wake time, and treat the diary like brushing your teeth — a brief routine, not a meticulous record.

• Estimate, don't measure. "I think I went to bed around 11:15" is fine. Don't lie awake with a stopwatch — that defeats the purpose. The diary is meant to capture habit, not stopwatch-precision data.
• Time to fall asleep. Use a rough estimate in minutes. If you don't remember, leave it blank rather than guess wildly — a blank field is more honest than an invented number.
• Wake-ups during the night. Only count wake-ups you actually remember. Brief micro-arousals you don't recall don't show up in conscious memory and shouldn't be invented.
• Quality rating. The 1–5 scale is intentionally coarse. A 3 means "okay, neither great nor awful" — most nights will be 3s and that's fine. A 1 or a 5 should mean something. If every night gets a 3, the rating isn't doing useful work; recalibrate.
• Lifestyle fields. Note actual amounts and timing, not vague "yes/no." "2 cups, last at 2 PM" is more useful than "yes." Exercise: type and time roughly — "30-min walk after work" is enough. If a field doesn't apply, leave it blank.
• Skip a day if you have to. A blank row is far better than a fabricated one. Two weeks of mostly-honest data is more useful than two weeks of mostly-imagined data.
• Use “Copy from yesterday” when your routine is steady. If today is mostly a repeat of yesterday — same bedtime, same caffeine routine, same alcohol intake — copy yesterday’s entries with one click and edit only the fields that actually changed (sleep onset, wake-ups, quality). Faster, and more accurate than retyping the same details from memory.

Consistency of when you fill it in matters more than the level of detail in any single entry. Twelve out of fourteen days completed on time will be far more useful than fourteen days of retrospective best-guesses written up on day fifteen.

What to do with the diary when it's done — for a clinician, or for yourself

When the two weeks are complete, the diary is most useful as a printed handout (or PDF) — at the start of a clinician appointment if you're scheduling one, or as your own reference document if you're tracking for personal reasons. The Print / Save as PDF button strips the page chrome and lays out all 14 days compactly across one or two pages. The Export CSV button gives you a spreadsheet-ready file if you want to compute averages yourself or import into another tool.

Whether you're sharing it with a clinician or reading it yourself, three things are usually worth looking at:

• Average total sleep time across the two weeks, and the spread (a 6-hour average with all nights between 5.5 and 6.5 reads very differently from a 6-hour average split between 4-hour nights and 8-hour nights).
• Consistency of sleep timing — does the bedtime/wake-up window land in roughly the same place every night, or shift by 90 minutes or more between weekdays and weekends?
• Sleep onset latency and wake-up frequency — flagged when onset regularly exceeds 30 minutes or wake-ups regularly exceed two per night, both of which point toward chronic insomnia or a comorbid disorder.

If you're bringing it to a clinician and the diary surfaces a pattern that meets one of those flags, bring it as data — not as a self-diagnosis. "I've tracked the last two weeks and I'm averaging 5.2 hours with a 45-minute onset on most nights" is the right framing. The clinician then decides whether the next step is a behavioral protocol, a referral for a home sleep apnea test, an in-lab study, or something else.

If you're using the diary for your own reasons — maybe you've been curious about your sleep for a while, or you want a baseline before trying a behavioral change — the same three signals are worth attention. A regular bedtime and wake time, consistent enough that the diary's two-week window shows little drift, is the foundation almost everything else builds on. Try a single behavioral change at a time (consistent wake time, caffeine cutoff, evening light reduction) for the next two weeks and run the diary again to see what moved.

When a diary alone isn't enough

A sleep diary is excellent at characterizing patterns the patient is aware of. It is poor at detecting events the patient sleeps through — and several of the most consequential sleep disorders fall into exactly that category.

• Sleep apnea. Patients with even severe obstructive sleep apnea often report sleeping seven or eight hours and don't remember the breathing pauses. The bed partner usually does. If a partner has reported loud snoring, witnessed apneas, or gasping awakenings — or you wake unrefreshed regardless of total sleep time — the diary will look unremarkable while the underlying problem is severe. See sleep apnea and home sleep apnea testing.
• Restless legs syndrome and periodic limb movements. The leg sensations that delay falling asleep show up in the diary; the periodic limb movements during sleep that fragment it usually don't. See restless legs syndrome.
• Narcolepsy and central hypersomnias. Daytime sleepiness in the face of apparently adequate nighttime sleep is the diagnostic signal, but the diary records the night, not the day. See narcolepsy.
• Parasomnias and REM sleep behavior disorder. Behaviors during sleep are observed by partners, not the patient. The diary at most records a vague awareness that something happened.

If you suspect any of the above, start with the Sleep Score — a 7-question screener that surfaces likely apnea risk, insomnia severity, and a composite sleep-wellness score, and points toward the right next step. The diary is a good companion to that screener, not a substitute. Use both, take both to a clinician (or keep both for yourself if you’d rather track first and decide later), and let the process unfold from there. Find a sleep specialist near you.

Privacy: how the diary is built

The privacy posture isn't a marketing sentence. It's an architectural commitment built into the page itself, and it's worth being explicit about what it means and what it doesn't.

• Diary content never leaves your device. Every entry is saved to your browser's local storage. Our server doesn't see it, our analytics doesn't see it, no third-party script sees it. There is no API endpoint receiving diary data because there is no API endpoint at all.
• There is no account. You don't log in. We don't know who you are. Closing the tab keeps your data; opening the page on a different device starts a fresh diary.
• There is no sync. A diary started on your phone is not visible on your laptop. This is a tradeoff — sync would require a server, and a server would mean diary content leaving your device, which is exactly what we're trying to avoid.
• There is no recovery. If you clear your browser data, switch to private/incognito browsing, or use the Clear all entries button, the diary is gone. We have no copy. This is what "no server storage" actually means.
• Page-load tracking is anonymous and aggregate. We use Plausible Analytics, which records that someone visited /sleep-diary/ and roughly which country they came from. It does not see diary content, does not set cookies, and does not build a per-user profile.

The button-click events fired by the action toolbar (start, print, CSV export, clear) record only the intent, never the data. The CSV export runs entirely in your browser using a Blob URL — the file is generated client-side and downloaded to your device without ever transiting our servers.

A sleep diary is a sensitive document. It records your sleep schedule, your alcohol consumption, your medication context, and frequently your stress and health concerns in the notes field. Treating it like the personal record it is — and architecting the tool so we couldn't read it even if we wanted to — is the only posture we think is defensible for an educational sleep platform.