How many calories break a fast?
This is the most searched question on the topic and it does not have the crisp answer people are hoping for. There is no officially agreed threshold. Commonly repeated figures include zero, ten, twenty and fifty calories, and each is stated somewhere as though it were established. No trial has determined a number that applies to everyone, and the question itself is harder than it sounds because different reasons for fasting imply different thresholds.
What we can be precise about is how this site treats it, because that is a definition rather than a discovery. For scheduling purposes a fast ends when you consume anything with meaningful calories, and the clock restarts from that moment. Plain water, black coffee and unsweetened tea are treated as compatible. That rule is applied identically everywhere on the site, so your history stays internally consistent.
If you want a practical rule to carry around, "if it has calories worth counting, it counts" gets you most of the way and is easy to apply at 11 PM without a reference table. The items in the trace band are the only genuine edge cases, and they are edge cases precisely because reasonable people put them on different sides of the line.
What can you eat during a fast?
Nothing, and that is not a trick answer. A fast is defined by not eating, so the question people mean when they search this is usually one of two others: what may I drink without ending the fast, or what should I eat once the eating window opens.
The first is the table above. Plain water, black coffee and unsweetened tea pass under every common definition; the trace band is genuinely disputed; everything with meaningful calories ends the fast and restarts the clock.
The second is a different page. What to eat when the fast ends, what to avoid, and why most of the advice online was written for multi-day fasts is covered in how to break a fast. Short version: for a daily fast, eat a normal meal.
One thing worth stating plainly, because the phrasing of these searches suggests real worry behind them: there is no food you are supposed to eat during a fasting window. If you are hungry enough that the question is pressing, that is useful information about whether the schedule fits you, and shortening the fast is a legitimate response rather than a failure.
What can you drink while fasting?
This is the question most people actually mean, and it has a short answer with a longer footnote. Water, black coffee and plain tea contain effectively nothing and pass under every common definition of a fast. Everything else depends on what is in it and how much you added.
The clear cases
Still and sparkling water, black coffee and unsweetened tea are the uncontested three. Nobody argues about them, and nothing in the table above puts them in question. If you want a rule you never have to think about again, that is the rule.
Where it gets argued about
A splash of milk, a diet soda, an electrolyte tablet, a squeeze of lemon. These sit in the band where opinions are confident and evidence is not. The measurable facts are in the table: a tablespoon of milk is a handful of calories, most diet sodas are zero, an electrolyte sachet varies enormously by brand. What nobody can tell you is the exact point at which those calories stop counting as nothing, because no trial has established one.
Our advice is unglamorous and it is the only defensible kind: pick where you draw the line and then draw it in the same place every day. A record built on one definition is comparable with itself, which is what makes it worth keeping.
What does not break a fast
Reading the list from the other direction is worth doing, because the anxious version of this question assumes the answer is short and it is not. Under any common definition, none of the following ends a fast: water in any form, black coffee, plain tea, most medications taken with water, brushing your teeth, chewing sugar-free gum by most people's reckoning, exercise, sleep, and being hungry.
That last one is on the list deliberately. Hunger is not a signal that something has gone wrong with the fast, and it is not evidence that you need to eat something small to keep the fast valid. There is no such mechanism. Sleep is on the list for the same reason: the hours count, which is why a 16-hour fast is mostly hours you were unconscious for.
What does end a fast, for scheduling purposes, is your first real bite. That is the moment the timer records and the boundary every calculation on this site works from.
Oils, and why they come up so often
Olive oil is the single most searched item in this whole category, well ahead of anything else, and MCT oil, coconut oil and oil in coffee follow behind it. The volume is a bit odd until you notice where it comes from.
The facts are simple. Oil is close to pure fat and carries roughly 120 calories a tablespoon, which is more than almost anything else people ask about here. By any threshold anyone has proposed, a tablespoon of olive oil is food. A few drops on the way past are a few calories, which puts them in the same disputed trace band as a splash of milk.
The reason it gets asked so much is that a strand of fasting advice suggests fat does not count because it does not move insulin much. That reasoning treats one hormone as the definition of a fast, which is a choice somebody made rather than a finding. If your definition is calories, oil breaks a fast and it is not close. If your definition is something else, you have adopted a definition that no trial has validated, and that is worth knowing rather than being talked out of.
Why consistency matters more than the threshold
Here is the part most pages skip. Whether you count a splash of milk as breaking your fast matters far less than whether you count it the same way every week. Your fasting history is a record you compare against itself, and changing the rule mid-stream makes January and March incomparable in a way no amount of precision elsewhere can fix.
This is the one part of the question genuinely within your control. The calorie threshold is unsettled and you cannot resolve it by reading more articles, because the articles disagree with each other. Your own consistency is a decision you can simply make. Pick a position on the trace items, write it down if it helps, and stop relitigating it every time you want a coffee.
Coffee, in more detail
Coffee generates more of these questions than everything else combined, so it is worth separating the drink from what goes in it. Plain black coffee is roughly two calories per cup, which is trace by any measure, and it passes under every common definition.
What people add is where it changes. A tablespoon of milk is around nine calories. Cream is fifty or more. Butter and MCT oil, which are sometimes marketed specifically as fasting-compatible additions, run to a hundred calories or more per serving and are among the most calorie-dense things you could put in a cup. Marketing language does not change the calorie count, and a drink built to be filling is, by construction, food.
Sweeteners are the genuinely disputed case. Zero-calorie sweeteners contribute nothing under a calorie-based rule, and the evidence on whether they matter for other reasons is mixed rather than settled. We are not going to resolve that for you, and any page that claims to has gone beyond what the research supports.
If you break your fast early
Recalculate and carry on. That is the whole procedure. Enter the actual time you ate into the break-fast calculator and it gives you the new schedule. If the timer is running, edit the start time and it recalculates.
Your history records what actually happened rather than what you planned, and it does so without failure language, warning colours or a broken streak. A 13 hour fast is recorded as 13 hours. Streaks exist only if you turn on Motivation Mode yourself, because for some people they help and for others they are exactly the wrong pressure. The default is off.
What this page will not tell you
You will find pages arguing that a specific number of calories does or does not interrupt autophagy at a specific hour. We are not going to join in. Those claims are extrapolated well beyond what human research supports, and comparing a few of them shows they disagree with each other, often substantially, while each states its number as fact. That disagreement is the useful signal.
What this page offers instead is the calorie content of each item, a plain verdict under a calorie-based rule, and honesty about which cases are genuinely disputed. That is less exciting than a metabolic timeline and considerably more likely to be true.