Which shift do you work?

Yes. You enter your hours and get your lowest point, your crash window, your sleep window and one of the timings without paying or making an account. The full schedule with all six timings, the day-by-day adaptation and the days-off protocol is the paid part.

A normal sleep calculator counts 90 minute cycles back from a bedtime you choose. It has no idea when you work. This one starts from where your body clock actually sits, positions your sleep against the hours your roster leaves free, and tells you when light will help and when it will undo the week.

Enter your own times. The presets cover the common blocks because most shift work runs on them, but the calculator takes any start and end time, any shift length, and any commute.

No. It runs in the browser and there is no account. If you want reminders at each timing you can add it to your home screen and turn them on, which is a phone setting rather than an app store download.

It is an estimate from your wake time, not a measurement. The only way to know your temperature minimum exactly is to measure it. If the number it gives you does not match the hour you actually feel worst, move it on the result and every other timing follows. The methodology page says which two constants are our judgement rather than the literature's.

There is no single healthiest roster, and anyone selling you one is guessing. Two things are consistently supported. Rotating forwards, days to evenings to nights, is easier than rotating backwards, because your clock moves later far more readily than earlier. And a longer run of the same shift beats constant switching, because it gives your clock time to arrive somewhere before it is moved again. Beyond that, the roster that works is the one whose sleep you can actually protect, which is why this tool asks about your commute and your days off rather than ranking schedules.

Aim to get your main sleep and any pre-shift nap adding up to what you would normally need over the whole 24 hours, not to hit a number in one block. In practice that usually means six or seven hours of day sleep plus a nap ending about an hour before you clock on. Day sleep runs shorter and breaks up more than night sleep, so two blocks you actually get beat one long one you keep losing.

It is a screening question, not a schedule. The clinical version asks whether disturbed sleep has run at least three nights a week, for at least three months, with at least three consequences during the day. It is routinely confused online with wind-down countdowns like 10-3-2-1, which are unrelated. As a self-test for a shift worker it is nearly useless, because the roster on its own ticks all three boxes without telling you what to change. If you think something is genuinely wrong, that is a clinician's question. What is adjustable is where your sleep sits relative to your worst hour, and that is what this computes.

A wind-down countdown, more often written 10-3-2-1-0: no caffeine ten hours before bed, no heavy food or alcohol three hours before, no work two hours before, no screens one hour before, no snooze button. None of it is wrong and all of it is anchored to a bedtime, which is the number a shift worker does not have. Fix the sleep window first and the countdown becomes usable. The caffeine figure is the one to adjust: ten hours is a blunt default, and a half life of roughly five hours puts the real cutoff nearer six hours before you sleep.

One main sleep block starting soon after you get home, a nap ending shortly before you clock on, sunglasses from the door of the building until you are in a blacked out room, and a last caffeine about six hours before the main block. That is the shape and it holds for most night rosters. The clock times behind it are not transferable: they are set by your finish time, your commute, and how early you would wake with no alarm. Enter those and you get six times instead of a shape.

We will not answer that, and you should be wary of a scheduling tool that does. The research linking shift work to long-term health outcomes is population-level association, and converting a population association into a prediction about one person is the step that turns a timing tool into a health claim it cannot support. What is measurable and what you can move is how much sleep you actually get and where your worst hour falls relative to your shift. Those two are what this works on. Anything about your own long-term health is for a clinician who can look at you.