How Many Ounces Should a Newborn Eat?

Search this question and you will get a dozen tables, each with slightly different figures, each stated with total confidence. That is not a sign that some of them are wrong and one is right. It is a sign that the underlying thing being described is genuinely variable, and that a table is a bad instrument for describing variable things.

Intake in the newborn weeks is not a fixed requirement that a baby either meets or misses. It moves with weight, with age in days, with whether a feed is at the breast or from a bottle, with how concentrated the feed is, with how long the baby slept before it, and with whether the baby is in the middle of one of those stretches where they seem to want everything, all night, for three days, and then settle again.

So a table has to choose. It can be narrow and useful for the average baby and wrong for yours, or it can be wide enough to include yours and too wide to tell you anything. Most choose narrow, because narrow looks authoritative. That is how you end up at 3am comparing your baby to a number that was never about your baby.

It is worth knowing the shape of official advice, because once you have seen the shape, the confident tables stop being intimidating.

The major pediatric bodies — the American Academy of Pediatrics in the United States, the World Health Organization internationally, the NHS in the United Kingdom — all publish material on infant feeding. Read it directly and you will notice what it has in common: it tends to be expressed as a range rather than a target, it is usually scaled to the baby's weight rather than stated as a flat figure, it changes with age, and it is almost always accompanied by an instruction to feed in response to the baby rather than to the number.

That last part is the part the tables leave out, and it is the most important sentence in any of these documents. The published ranges exist to describe a population, so that a clinician looking at a whole caseload has something to orient against. They were never intended as a score for one family to grade one baby against at 3am.

We are deliberately not reproducing figures here. Any number we printed would be stripped of the weight, the age, and the clinical context that make it meaningful, and you would remember the number and forget the context. If you want the numbers, get them from your pediatrician, who can attach them to your actual baby.

It helps to know why the variation exists, because it makes your own baby's pattern feel less like a deviation and more like a fact about them.

Ask a pediatrician what they want to know about feeding, and they will rarely ask you to compare your baby to a chart. They will ask about your baby, over time: roughly how much in a day, roughly how many feeds, how that compares to a couple of weeks ago, and whether anything changed suddenly.

That is a different kind of number, and it has a property no chart can have. It is a baseline. Once you know what an ordinary week looks like for your baby, you can recognize a change, and a change is the thing worth mentioning. One feed on its own is hard to read; a change across several days is what's worth describing to your pediatrician. A week that is consistently different from the week before is a real observation, and it is one you can only make if you wrote the weeks down.

This is the practical answer to the question in the title. You cannot know how many ounces your newborn should eat. You can know how many ounces your newborn does eat, on an ordinary week, and you can notice when that shifts. The second is more useful than the first, and it is entirely within your reach.

You do not need a lot. A useful feeding record is smaller than you might assume, and the surplus detail is what makes tracking collapse by week two.

Three things carry almost all of the value.

There is a way to look at your own numbers that is genuinely informative, and a way that is just anxiety with a chart attached. The difference is whether you are comparing your baby to themselves or to an imagined standard.

Useful questions sound like this. Is the daily total roughly where it was last week, higher, or lower? Are feeds clustering at a particular part of the day? Has the gap between feeds stretched out, and did that start on a particular day? Did anything change around the time the pattern changed — a new bottle, a cold, a trip, a different person doing the evenings?

Unhelpful questions sound like this. Is this number good. Is my baby behind. Is this enough. Those questions cannot be answered by a log, they cannot be answered by a chart, and at 3am they cannot be answered by you. They are questions for a person who has examined your baby.

This is also why we think tracking apps should not flag numbers. An app that has never met your baby is in no position to call a feed low, and 3am is the hour you are least able to shrug off a red badge.

Sometimes the search is not about arithmetic. Sometimes it is about a feeling that something is off, and ounces are just the nearest measurable thing to grab.

If that is where you are, the log is not the tool. Anything that worries you is reason enough to call. Call your pediatrician's office: they field these calls constantly, including at night, including for things that turn out to be nothing, and they would far rather hear from you early than have you sit with it until Thursday. Feeding changes, unusual sleepiness, difficulty waking for feeds, changes in how many wet or dirty diapers you are seeing, or anything about how your baby seems in themselves are all reasonable reasons to pick up the phone. In an emergency, call 911.

What the log does is make that call better. Instead of “I think she's been eating less,” you can say what you actually saw, over how many days, compared to what. That turns a worry into information, and information is what a clinician can work with.