How Much Milk Should I Pump? Daily and Per-Session Numbers

Most pumping charts give you one number, and it’s rarely the number you need. What you should pump depends on one question: is this session replacing a feed, or is it extra? This guide gives you the measured figures behind the usual answers, the math to apply them, and an honest look at what the studies can’t tell you. All the per-feed amounts by age are in our calculator, and our pumping hub covers the rest of the basics.

The daily total is the number that matters

Start with what a day of milk looks like, because sessions vary and days are steadier.

In a 2006 study at the University of Western Australia, Jacqueline Kent and colleagues had 71 mothers of exclusively breastfed babies aged 1 to 6 months weigh their babies before and after every feed for 24 to 26 hours. As a 2021 University of Western Australia review reports that study, 24-hour intake averaged 788 ± 169 g (about 27 oz), with a range from 478 to 1,356 g (16 to 46 oz). So one mother’s normal day can be almost three times another’s. Babies fed 11 ± 3 times a day, and each feed from one breast averaged 76 g (about 2.6 oz).

A later 2013 study from the same university followed mothers over time and found that 24-hour intake stayed constant from 1 to 6 months. In a 2025 University of Western Australia study of 281 mothers measured the same way, the mean was very similar at 778 ± 238 g.

That’s why your daily goal doesn’t keep climbing the way formula charts do. To convert, we treat 1 g of milk as about 1 ml and 1 oz as about 29.6 ml. The papers use a density of 1.03 g/ml, so these conversions are slightly rounded.

What different sources call “enough”

There’s no single official target, and the sources below were written for different people. This comparison is our own.

Source	Figure	Who it applies to
ABM Protocol #35 (2021)	700 to 980 ml (25 to 35 oz) a day by days 10 to 14, kept up for 6 months	Goal for people separated from a baby and fully replacing nursing
Kent 2006	Mean 788 g; range 478 to 1,356 g, as reported in the 2021 review above	Measured, not a goal: exclusively breastfed, 1 to 6 months
Gridneva 2025	Under 600 ml a day classed as low	Research cutoff, from 2 weeks to 6 months
Kent 2016	440 ml a day as the lower limit of normal, by day 11	Term babies, early weeks
Hill 2005, University of Illinois at Chicago	At least 500 ml a day at week 6 counted as adequate	95 pumping mothers of preterm babies and 98 mothers nursing term babies
Mago-Shah 2023, Duke	More than 500 ml a day by day 14 (“coming to volume”)	NICU mothers; of 427 who produced milk, about half reached it

The disagreement that matters: a 2026 University of Western Australia paper notes published thresholds run from 500 ml a day by day 14 to 750 ml by day 10. As we read it, some of that gap comes from purpose: the lower numbers flag who needs help, and the higher ones are a goal for feeding a baby entirely on pumped milk. Falling between them isn’t a failure. It’s a reason to compare your total with your baby’s actual intake.

The per-session math

Daily total ÷ number of sessions = average per session (both breasts combined). These calculations are ours, using the study figures above.

Daily total	5 sessions	6 sessions	8 sessions	10 sessions
600 ml (20 oz), low cutoff	4.1 oz	3.4 oz	2.5 oz	2.0 oz
788 ml (27 oz), Kent mean	5.3 oz	4.4 oz	3.3 oz	2.7 oz
1,000 ml (34 oz), near the top of ABM’s goal	6.8 oz	5.6 oz	4.2 oz	3.4 oz

How does that compare with real sessions? In a 2026 University of Western Australia crossover trial of 28 mothers, about 3.5 months postpartum on average, one session yielded about 147 g on a hospital-grade pump and 125 g on a personal-use pump from both breasts combined (roughly 4 to 5 oz). Spread was wide, at about ± 75 g. After the researchers adjusted for how full the breast was beforehand, the pumps did not differ. Both came from one manufacturer, and the authors say the result can’t be generalized to other makers’ pumps.

Worked examples

These are illustrations using average figures, so yours will differ.

• Exclusively pumping, 8 sessions a day. Say your baby drinks the Kent average of 788 ml (about 27 oz). You’d want about 3.3 oz per session on average. A 2 oz session at bedtime and a 5 oz session at dawn can still add up, so log the total, not each bottle.
• Pumping at work, away 9 hours. Kent’s mean is about 33 g an hour, or 1.1 oz (788 ÷ 24). Nine hours is about 10 oz, so three sessions of about 3.3 oz would match. Our calculator’s typical range for 9 hours is about 7 to 11 oz.
• Nursing on both sides, then pumping right after. Kent found babies took about 67% of the milk available. If 76 g was 67%, about 113 g was available and about 37 g stayed behind per breast. If your pump then removes about two-thirds of that, you’d collect about 25 g per breast, or about 1.7 oz from both sides. An hour later, with about 33 g of new milk, it’s closer to 2.4 oz. This is a rough model. It shows why 1 to 2 oz after a feed tells you very little about your supply.

Output by age

Stage	What was measured	Notes
Day 2	Median 22 ml in 24 hours	Perrella 2026: 44 mothers after birth at 28 to 34 weeks (30 with Day 2 data), expressing about 6 to 7 times a day
Day 6	Median 532 ml (middle half 217 to 685 ml)	Same study (38 with Day 6 data); 61.5% of those with Day 13 data had reached 600 ml by then
Days 6 to 28	Mean 693 ml transferred by nursing	Kent 2016: only 13 term mothers without problems; those with problems were lower
1 to 6 months	Mean 778 to 788 g; no rise with age	Kent 2006, 2013; Gridneva 2025
6 to 12 months	Varies with solids	ABM Protocol #35 gives about 800 ml (450 to 1,200 ml) for exclusive breastfeeding at 6 months

We found no study that measured output by age specifically in term mothers pumping from birth, and that’s a real gap. A 2026 University of Rhode Island scoping review of expression in mothers of full-term babies reported that women lack practical guidance, especially those exclusively pumping. So treat the Day 2 and Day 6 figures as a rough guide. Those mothers delivered early, and term figures may differ. Solids change the picture after 6 months, because babies drink less milk as they eat more food.

Does a pump really get less than a baby? A myth check

You’ll often read that a baby always removes more milk than a pump, so your pump number is an underestimate. The measured data are more interesting than that.

Group	Percentage of available milk removed
Babies nursing (Kent 2006)	67.3% ± 7.8
Pump, 15 minutes at maximum comfortable vacuum (Kent 2008, 23 mothers)	65.5% ± 4.1
Occasional pumpers, 24-hour average (Gridneva 2025, 58 mothers with this measure)	70.0% ± 10.9
Nursing only, 24-hour average (same study, 114 mothers with this measure)	75.8% ± 13.2
Exclusive pumpers, 24-hour average (same study, 11 mothers with this measure)	83.0% ± 8.4

So a good pump, used at a comfortable strong setting, can remove about the same share as a baby. The University of Western Australia authors call the usual pump-versus-baby comparison incongruous, because a baby feeds to appetite rather than until the breast is empty.

Limits worth knowing: this research comes largely from one university group, supported by an unrestricted grant from the pump maker Medela (the authors say the funder had no role). Many sessions used a hospital-grade pump and fitted shields under controlled conditions, and exclusive pumpers are a small, select group. Older work from Georgetown University in 23 women found that pumps differed strikingly in prolactin response. Your particular pump may not match these numbers.

Why your number can still read low

• You’re measuring leftovers. See the worked example above.
• Vacuum and fit. In Kent 2008, softer vacuums yielded less milk. Simultaneous pumping removed more than one side at a time in a 2012 University of Western Australia study of 31 mothers. A flange that doesn’t fit can also make pumping uncomfortable, so recheck your size if your output or comfort changes.
• Letdown at work or under pressure. ABM Protocol #35 says images, sounds or smells of your child can help trigger letdown, and warmth, gentle massage and music may help volume. It also says mothers may need reassurance if pumped volumes look low.
• Normal unevenness. Kent 2006 found volume varied by time of day and by breast, so one low session isn’t a trend.

Exclusive pumping: what the numbers show

This is the best recent data, and one detail tends to get lost.

A 2025 University of Western Australia study looked at 93 mothers who predominantly pumped. The 32 exclusive pumpers in its normal-supply group produced 1,018 ± 274 g a day (34 oz), more than the 792 ± 219 g in a comparison group that nursed or pumped occasionally. They averaged 6.2 ± 1.8 sessions a day, and their babies drank 773 ± 184 g of expressed milk.

Our calculations from those figures: 1,010 g expressed over 6.2 sessions is about 163 g, or 5.5 oz, per session. Production was also roughly 245 g (about 8 oz) above what babies drank, though these averages come from slightly different subgroups.

What the headline doesn’t show: the paper says all exclusive pumpers kept production above 600 g. But that figure is for the group sorted by normal supply. In the low-supply group, 9 of 26 mothers also weren’t breastfeeding at all (the paper lists 17 who were), so by our count 32 of 41 (78%) mothers who only pumped reached 600 g. In the university’s later study, exclusive pumpers ranged from 158 to 1,739 g. Mothers also came to the study already pumping, between 2 weeks and 6 months after birth, so those who stopped early due to low supply wouldn’t be counted. These studies show that exclusive pumping can sustain a full supply. They don’t show that it will for everyone.

Six sessions shouldn’t be read as a target for the early weeks either. In Duke University’s NICU study, reaching 5 or more sessions a day by day 5, and pumping between 1 and 5 a.m., were each independently linked with reaching 500 ml by day 14. That’s an association, not proof of cause. Our exclusive pumping guide and pumping schedule cover how to set up your day, usually with more sessions at first.

More isn’t always better

Once you’re meeting your baby’s needs, chasing a higher number has a cost. ABM Protocol #36 says, in its guidance on mastitis, that “pumping to empty” perpetuates a cycle of hyperlactation and is a major risk factor for worsening tissue swelling and inflammation, and that people using pumps should express only the volume their baby consumes. If you’re pumping far above your baby’s intake and getting repeated pain or lumps, raise it with your clinician.

How to use your own numbers

This decision guide is ours, not a clinical protocol.

1. Add up one full day, all sessions plus any nursing you can’t measure. Don’t judge from one session.
2. Compare it with your baby, not with a chart. A baby who is gaining well is getting enough, whatever a pumping forum says.
3. Nursing most feeds? Your pump number mostly tells you how much you have stored for bottles. Plan on what you need for time apart.
4. Replacing all or most feeds? Compare your total with your baby’s bottles. If you’re consistently short, look at session count, fit and vacuum before adding supplements. Our guide on how to increase milk supply has the steps.
5. Sudden fall in a steady number? Read our guide to a sudden drop in milk supply to narrow down the cause.

When to call a doctor or lactation consultant

Call your pediatrician if your baby seems very sleepy, is hard to wake for feeds, has fewer wet diapers than expected (the AAP looks for 6 or more a day by 5 to 7 days old), or isn’t gaining weight.

Talk to an IBCLC if your total stays below about 20 oz (600 ml) a day after the first 2 weeks while you’re relying on pumped milk, if pumping hurts, or if your output falls for several days in a row. These cutoffs come from research and don’t diagnose anything, but they’re a good reason to get your pump and routine checked. Call your doctor if you develop a hot, red, painful area on your breast or a fever.