Understanding Milk Production: The Hormonal and Physiological Science of Lactation

By Birth & Baby University · Last updated: October 2026 · 8 min read

Few topics generate more worry in the newborn weeks than milk supply. Parents compare feeding times, pump output and the weight of a baby against numbers from friends or the internet. A postpartum doula who understands the basic science of lactation can answer the most common worries accurately: how milk “comes in,” what drives production, why breast capacity and output vary, and which patterns are within the range seen in research. This article summarizes the physiology as described in reviews and studies, and flags where the literature is limited or where confident-sounding claims run ahead of it.

Key takeaways

Quick answer: milk production is set in motion by hormonal changes after birth and then largely governed by how much milk is removed from the breast. The hormones prolactin and oxytocin both play central roles, and stress or some medical interventions can blunt oxytocin release. Normal feeding patterns vary a great deal from one mother-baby pair to the next.

• A physiology review describes copious milk secretion (lactogenesis stage II) as triggered by a prepared mammary gland, progesterone withdrawal, maintained prolactin and removal of milk from the breast soon after birth.
• A review of how breastfeeding works says the amount of milk produced depends on the amount removed, and that exclusively breastfed babies show a three-fold variation in the amount of milk they take per day.
• In a study of 71 mothers, babies fed 11 times a day on average with a range of 6 to 18, and there was no relationship between the number of feeds and the 24-hour milk production.
• In a cohort of 431 first-time mothers, 44% had delayed onset of copious milk (after 72 hours), and independent risk factors included maternal age 30 or older, overweight or obese body mass index, and an infant who did not “breastfeed well” at least twice in the first 24 hours.

What the research says

Starting milk production. Neville, Morton and Umemura’s review of lactogenesis (2001) describes the shift from the quiet mammary gland of pregnancy to active secretion. In its summary, the most important factors in starting copious milk secretion (lactogenesis stage II) appear to be a prepared mammary epithelium, a fall in progesterone after delivery, maintained prolactin and removal of milk from the breast within an undefined interval after birth. The authors acknowledge that the molecular details of how progesterone and milk removal act on mammary cells were not well studied at the time, and that useful clinical studies were still needed. Practically, this is the physiological basis for the advice to feed or express milk early and often.

Oxytocin and prolactin. Uvnäs Moberg and colleagues’ 2020 systematic review of maternal oxytocin during breastfeeding included 29 articles covering 601 women. It found that breastfeeding triggers an immediate oxytocin release lasting about 20 minutes, pulsatile early postpartum (about 5 pulses per 10 minutes) and becoming a more protracted rise as lactation proceeded. The number of pulses during early breastfeeding was associated with greater milk yield and longer duration of lactation, and was reduced by stress. Oxytocin release was associated with higher prolactin and lower stress hormones. The authors reported that mechanical pumping, but not bottle-feeding, was associated with oxytocin and prolactin release; that emergency cesarean birth reduced the oxytocin and prolactin response to breastfeeding; that epidural analgesia reduced prolactin; and that synthetic oxytocin infusions increased prolactin. All studies were exploratory in design, so the review is a map of associations, not proof of effects.

Removal drives production, and patterns vary. Kent’s 2007 review in the Journal of Midwifery & Women’s Health concludes that the amount of milk produced depends on the amount removed, that the amount and composition of milk are largely independent of the mother’s diet, and that if a baby is growing normally the mother can be confident that the baby does not need to follow prescribed regimes. She should respond to the baby’s cues about frequency and whether one or both breasts are needed for a meal. The supporting data come from Kent and colleagues (2006), who test-weighed 71 exclusively breastfeeding babies aged 1 to 6 months at every feed for 24 to 26 hours. Babies fed 11 plus or minus 3 times in 24 hours (range 6 to 18), and the average feed was 76 grams, about two thirds of the milk available in the breast at the start. The two breasts rarely produced the same volume, night feeds made an important contribution, fat content was independent of feeding frequency, and there was no relationship between the number of feeds per day and the mother’s 24-hour milk production. The authors conclude that babies should be encouraged to feed on demand, day and night, rather than conform to an average. The studied mothers were already breastfeeding well at 1 to 6 months, so these findings describe established lactation, not the first days.

When milk is slow to come in. Nommsen-Rivers and colleagues’ study of 431 first-time mothers (2010) defined delayed onset of lactogenesis as occurring after 72 hours. The median onset was 68.9 hours and 44% of mothers were delayed. In a multivariable model, independent risk factors were maternal age 30 or older, overweight or obese body mass index, birth weight over 3600 grams, absence of nipple discomfort in the first 3 days and a baby failing to “breastfeed well” at least twice in the first 24 hours. The authors call the risk factors multidimensional and the problem alarmingly common among first-time mothers. These are associations in one cohort, and delayed onset by their definition is not the same as insufficient milk.

Perceived versus actual supply. Huang and colleagues’ 2022 systematic review of perceived insufficient milk supply included 27 studies. It found that about half of mothers who stopped breastfeeding gave perceived insufficient supply as a reason, while among mothers who were still breastfeeding, 10% to 25% reported it. Perceived insufficiency was strongly related to lower breastfeeding self-efficacy and moderately related to infant suck ability and formula supplementation, and infant crying was reported to be read as a sign of it. The review studies perception, not measured supply, which is why confidence-building and skilled assessment matter.

What this means for doulas

• Explain the engine. “Milk is made in response to milk being removed, so frequent, effective feeding early on matters” is consistent with the physiology reviews.
• Normalize variation. Feeding frequency ranged from 6 to 18 times a day in healthy dyads in one study, and the breasts differed in output. A single “right” schedule does not exist.
• Do not promise or blame. Stress and medical interventions are associated with reduced oxytocin release, but the evidence is exploratory. Avoid telling parents that they caused a supply problem by being anxious.
• Treat delayed milk as common. If milk is not in by day 3 or 4, many first-time parents are within the range seen in the cohort. It still deserves attention from the care team, particularly alongside weight loss, few diapers or a sleepy baby (see Newborn Weight Loss and Weight Gain: What’s Typical and When Is Evaluation Needed?).
• Refer for assessment. Judging milk transfer, diagnosing low supply and recommending galactagogues are clinical or lactation-professional tasks.

Putting the evidence into practice

1. Ask about removal. How often is the baby feeding or the parent expressing, and how comfortable is it? Early effective removal is the lever the physiology identifies.
2. Protect skin-to-skin time and rest. These support feeding and may reduce stress, though the research on direct effects on supply is limited.
3. Explore confidence. Because perceived insufficiency is tied to self-efficacy, ask what the parent has noticed and what worries them, then connect them to a lactation professional if intake is in question.
4. Use objective cues. Wet and BM diapers and weight checks with the clinician are better indicators than breast fullness or the baby’s fussiness.
5. Link concerns together. See How to Recognize Effective Milk Transfer Without Relying on Feeding Duration Alone and Cluster Feeding: What Is Normal and What Might Indicate a Feeding Concern?.

Limits and uncertainty

Much of the physiology evidence comes from small studies, including exploratory hormone studies and test-weighing studies of well-established breastfeeding, and some key work is older. The Kent data describe mothers whose babies were 1 to 6 months old and feeding on demand; they do not set targets for newborns. The delayed-onset cohort is observational, from one site and population. The perceived-supply review combines studies of different quality and measures. This article draws on abstracts rather than full texts, and the lactogenesis review abstract is brief. Nothing here is a protocol for treating low supply, and medications, endocrine conditions, breast surgery or anatomical issues require clinician guidance.

The bottom line

The science of lactation is largely a science of response: hormones start milk production, and removal of milk, with the right oxytocin-supporting conditions, keeps it going. Variation between mother-baby pairs is normal, delayed onset of copious milk is common, and worries about supply are widespread even when supply is adequate. A doula adds the most value by explaining the process, supporting early and responsive feeding, protecting parents’ confidence and referring early when intake is in question.

Birth & Baby University’s For Doulas page sets out its birth and postpartum doula training, recertification and mentorship for professionals who support families through feeding and the early weeks. Check that any program meets your certifier’s, employer’s or state’s requirements before you enroll.

This article is for professional education. It is not medical, legal or tax advice. Check the standards of your certifying organization and the rules where you work. For concerns about a baby’s health, contact a pediatric provider or emergency services.

Sources and further reading

• Neville MC, Morton J, Umemura S. Lactogenesis: the transition from pregnancy to lactation. Pediatric Clinics of North America (2001). Read the source
• Uvnäs Moberg K, et al. Maternal plasma levels of oxytocin during breastfeeding: a systematic review. PLOS ONE (2020). Read the source
• Kent JC. How breastfeeding works. Journal of Midwifery & Women’s Health (2007). Read the source
• Kent JC, et al. Volume and frequency of breastfeedings and fat content of breast milk throughout the day. Pediatrics (2006). Read the source
• Nommsen-Rivers LA, et al. Delayed onset of lactogenesis among first-time mothers is related to maternal obesity and factors associated with ineffective breastfeeding. American Journal of Clinical Nutrition (2010). Read the source
• Huang Y, et al. The rates and factors of perceived insufficient milk supply: a systematic review. Maternal & Child Nutrition (2022). Read the source

More for professionals: the Professionals Blog.