A client sends me the sleep diary the morning of her second session. Seven mornings of the same five numbers: time into bed, time trying to fall asleep, minutes to sleep onset, number of wake-ups, time out of bed. It is the most useful record of her sleep that exists, and it changes what happens in that second session more than anything she could tell me from memory.

The Sleep Diary Is the Treatment Plan, Not Homework Attached to It

Most people expect a sleep diary to be a warm-up exercise, something to fill in before the real therapy starts. It is the real therapy. CBT-I is not a fixed script delivered the same way to every client. It is a protocol tuned to what a specific person's sleep is doing, and that tuning happens through the diary. Without it, a clinician is working from memory and impression, both of which are unreliable for something as variable as sleep.

This distinction matters because it changes what tracking is for. Just like a food diary in a weight-loss program, the sleep diary increases awareness of patterns a person would otherwise never notice. But it also does something a food diary does not: it gives us the measurements to calculate the exact bedtime and rise time a client will use starting the following week. Skipping the diary does not mean skipping a formality. It means the sleep restriction prescription has nothing to be built on.

The Consensus Sleep Diary Standardizes What Gets Tracked

There have been many versions of the sleep diary over the decades, which created a problem: clinicians and researchers were not always measuring the same things the same way, making it hard to compare results across studies or even across sessions with the same client. In 2012, a panel of insomnia researchers led by Colleen Carney published the Consensus Sleep Diary¹, an expert-agreed format built and tested with focus groups of people with insomnia, good sleepers, and people with sleep apnea, then refined for plain-language clarity so it could be completed by almost anyone. It is the same format we use with clients throughout treatment.

The core version asks nine questions each morning: what time you got into bed, what time you tried to fall asleep, how long it took, how many times you woke during the night, how long those wake-ups lasted in total, your final wake-up time, what time you got out of bed, and a rating of sleep quality. That is the entire instrument. It takes under two minutes to complete and produces four numbers a clinician uses: total time in bed, total sleep time, sleep onset latency, and wake after sleep onset.

Sleep Restriction Doses Are Calculated From Diary Numbers

Sleep restriction, the component of CBT-I that does most of the heavy lifting, works by temporarily narrowing the window of time a person spends in bed so that the sleep drive concentrates into that window instead of spreading thin across a longer one. The starting window is not a guess and it is not the same for every client. It comes directly from the diary's average total sleep time across the baseline period.

Without diary data, there is no defensible way to set that starting window. Too generous and sleep drive stays diluted, which is the exact pattern that keeps insomnia going. Too restrictive and a client is sleep-deprived for no benefit. The diary is what keeps this calculation grounded in what is happening rather than what a generic protocol assumes about an average patient who does not exist.

The Diary Catches the Gap Between Feeling and Fact

A meaningful number of people with chronic insomnia experience what sleep researchers call sleep-wake state misperception, sometimes labelled paradoxical insomnia. Objective measures show they are sleeping a reasonable number of hours, but their subjective experience is of lying awake most of the night. A 2018 review in Sleep Medicine Reviews² describes this subjective-objective discrepancy as common in insomnia disorder and identifies conditioned hyperarousal as the leading explanatory model, though the mechanism is still not fully settled.

A sleep diary does not resolve this discrepancy by itself, since it is still a subjective report. What it does is make the pattern visible and trackable over time, which polysomnography in a single overnight lab visit cannot do. Several nights of consistent diary data let a clinician see whether a client's perceived sleep onset latency is drifting closer to what the behavioural interventions would predict, which is itself useful clinical information regardless of what a single night's actigraphy might show.

Wearables Are Not a Substitute

Clients often ask whether their Apple Watch, Oura Ring, or Fitbit can replace the diary, since the device already produces sleep duration and stage estimates automatically. The honest answer is no. A 2025 meta-analysis in the Journal of Clinical Sleep Medicine³ pooled data from 798 patients across 24 validation studies and found consumer wrist-worn trackers differ from polysomnography on every major sleep parameter: total sleep time, sleep efficiency, sleep onset latency, and wake after sleep onset all reached statistical significance, with total sleep time off by close to 17 minutes on average and sleep efficiency off by nearly 5 percentage points. A Canadian expert consensus on insomnia management⁴ reached the same conclusion independently, recommending that data from sleep trackers be used with caution when estimating sleep amount and quality.

The discrepancy tends to run in a specific direction for insomnia specifically. Devices that rely on movement sensing often mistake lying still and awake for sleep, which is precisely what many people with insomnia do for long stretches of the night. That is the opposite of reassuring for a population whose central problem is already a mismatch between how sleep feels and how it is measured.

There is also a documented downside to leaning on tracker data as a primary source of sleep information. Clinicians at Rush University coined the term orthosomnia in 2017⁵ to describe patients developing a perfectionistic preoccupation with optimizing their tracker-reported sleep scores, sometimes reporting more distress about a bad number on the app than about how they slept. For someone already anxious about sleep, a device delivering a nightly grade on performance can add fuel rather than clarity.

None of this means wearables are worthless or that the research conversation is closed. The prospect of pairing the diary's subjective account with reliable objective data is one researchers are actively pursuing, and the same 2017 paper that named orthosomnia also argued that incorporating consumer sleep technology into CBT-I thoughtfully will become increasingly important as device use grows.⁵ The current limitation is accuracy, not concept. Until validation studies close that gap, a wearable can sit alongside a sleep diary as a point of curiosity, but it should not replace it as the basis for a treatment decision.

Two Weeks of Baseline, Not Forever

Tracking in CBT-I is not indefinite. Before baseline tracking begins, your sleep and insomnia assessment establishes whether CBT-I is the right fit and rules out other sleep disorders. A baseline period of one to two weeks then establishes where sleep stands before any intervention begins, matching what Canadian clinical guidance recommends⁶ for insomnia assessment. Once treatment starts, diary entries continue on a nightly basis through the active phase, typically six to eight weeks, because each week's schedule adjustment depends on the previous week's numbers. After treatment ends, most clients stop daily tracking altogether and only pick the diary back up briefly if sleep starts to slip. You can read more about what that active phase looks like in our learning hub.

Some clients worry that watching their sleep this closely, every single morning, will make them more anxious about it, not less. That concern is reasonable. The distinction is between measurement and monitoring for perfection. A diary entry takes under two minutes, asks for rough estimates rather than precise clock-watching, and is completed once a day after waking, not throughout the night. It is structurally different from checking a phone screen mid-sleep or refreshing an app's sleep score. For most clients, the anxiety fades within the first week as the numbers start doing something useful instead of sitting there as evidence of a problem.