Cannabis use among breastfeeding mothers is increasing. A 2020 study published in Pediatrics by Garry et al. found that cannabis use among pregnant and postpartum women in the United States increased from 3.4% in 2002 to 7.0% in 2017. Among breastfeeding mothers specifically, survey data suggests that 5–8% report active cannabis use during lactation, with rates higher in states with legal adult-use markets.
This reality creates an urgent need for clear, evidence-based information. The standard medical advice — do not use cannabis while breastfeeding — is well-founded. But many parents want to understand why, what the actual risks are, and what the data shows versus what is assumed. Here is what the research has established.
THC Transfers Into Breast Milk: The Pharmacokinetics
The fundamental fact underlying all concerns about cannabis and breastfeeding is that THC readily transfers into breast milk. This is not speculative — it has been directly measured in multiple studies.
THC is highly lipophilic (fat-soluble), and breast milk is a lipid-rich fluid. The fat content of human breast milk ranges from 3–5%, providing an ideal medium for THC accumulation. THC concentrates in breast milk at levels that exceed plasma (blood) levels, with breast-milk-to-plasma ratios estimated at approximately 8:1 in some studies.
Key pharmacokinetic studies:
Baker et al. (2018), Obstetrics & Gynecology. This study directly measured THC concentrations in breast milk from 50 breastfeeding women who used cannabis. THC was detected in 63% of breast milk samples collected up to six days after last reported use. The mean THC concentration was 9.47 ng/mL. Notably, THC was detectable in samples collected from women who reported their last use as six or more days earlier, confirming the prolonged persistence of THC in breast milk.
Bertrand et al. (2018), Pediatrics. A carefully controlled study that collected 54 breast milk samples from 50 women who used cannabis. THC was detectable in 34 of 54 samples (63%). THC-COOH (the inactive metabolite) was detected in 5 samples. CBD was detected in 5 samples. The estimated infant daily dose of THC was calculated at 2.5% of the maternal dose — the same metric used for other drugs during lactation.
Moss et al. (2021), JAMA Pediatrics. This study examined the kinetics of THC in breast milk over time after a single episode of cannabis use. In 7 breastfeeding women who used a measured dose of inhaled cannabis, THC concentration in breast milk peaked at approximately 1 hour post-use and remained detectable for a median of 5 days (range: 4–6 days). The peak breast milk THC concentration was 94 ng/mL, declining over the following days.
| Study | Samples | THC Detection Rate | Key Finding |
|---|---|---|---|
| Baker 2018 | 50 women | 63% of samples | THC detectable up to 6+ days post-use |
| Bertrand 2018 | 54 samples | 63% of samples | Infant dose ~2.5% of maternal dose |
| Moss 2021 | 7 women, serial sampling | 100% at 1 hour | THC peaks at ~1 hour; detectable 4–6 days |
The “Pump and Dump” Question
Many breastfeeding parents ask whether they can use cannabis and then “pump and dump” — expressing and discarding breast milk for a period after use, then resuming breastfeeding once the milk is “clean.”
The pharmacokinetic data makes clear that pump-and-dump is not a reliable strategy for cannabis. Unlike alcohol, which clears from breast milk in a few hours as blood alcohol levels fall, THC persists in breast milk for days due to its lipophilic nature and sequestration in adipose tissue.
A 2018 analysis in Clinical Chemistry by Wymore et al. calculated that even with aggressive pumping, a single cannabis use event could produce detectable THC in breast milk for 4–6 days. For regular users, THC levels in breast milk may never fully clear between uses, resulting in chronic low-level infant exposure.
The fundamental problem is pharmacological: alcohol is water-soluble and clears proportionally with blood levels. THC is fat-soluble and accumulates in lipid-rich tissues (including mammary tissue), from which it slowly releases over days. There is no practical pump-and-dump window that ensures zero THC in breast milk after cannabis use.
What Does the Infant Actually Absorb?
Even though THC is present in breast milk, the critical question is: how much THC does the infant actually absorb, and does it reach biologically significant levels?
Oral bioavailability in infants. When an infant ingests THC via breast milk, it must be absorbed from the GI tract and survive first-pass hepatic metabolism before reaching systemic circulation. Oral THC bioavailability in adults is approximately 6–20%. Infant oral bioavailability is unknown but may differ due to immature hepatic metabolism.
Estimated infant exposure. Based on the Bertrand et al. data, an infant consuming 150 mL/kg/day of breast milk containing 9.47 ng/mL of THC would ingest approximately 1.42 mcg/kg/day of THC. For a 4 kg infant, this is approximately 5.7 mcg per day. With oral bioavailability of 6–20%, the systemically absorbed dose would be approximately 0.3–1.1 mcg.
For context, a typical adult recreational cannabis dose delivers 5,000–25,000 mcg (5–25 mg) of THC. The infant dose is roughly 1/5,000th to 1/25,000th of an adult dose. However, this comparison is misleading because infants have dramatically lower body weight, immature blood-brain barriers, developing endocannabinoid systems, and immature hepatic metabolism. A dose that is trivial for an adult may not be trivial for a developing brain.
THC has been detected in infant specimens. A 2016 case report in Clinical Chemistry documented detectable THC levels in the urine and stool of an exclusively breastfed infant whose mother used cannabis regularly. This confirms that ingested THC from breast milk does reach the infant’s systemic circulation.
Potential Effects on Infant Development
This is where the evidence becomes most limited and most concerning.
Endocannabinoid system in infant development. The endocannabinoid system plays critical roles in brain development during infancy and early childhood. CB1 receptors are expressed in the developing brain and are involved in neuronal migration, axonal pathfinding, synaptogenesis, and myelination. Exogenous THC exposure during this period has the theoretical potential to disrupt these processes.
Animal data. Rodent studies have consistently shown that postnatal THC exposure produces long-lasting neurodevelopmental changes. A 2017 study in Neuropharmacology by de Salas-Quiroga et al. found that perinatal THC exposure in mice altered glutamatergic neuron development in the cortex, produced seizure susceptibility, and caused cognitive deficits that persisted into adulthood. A 2019 study in Molecular Psychiatry found that perinatal cannabinoid exposure produced long-term changes in stress reactivity and social behavior in rats.
Human data. Direct evidence on the effects of cannabis-in-breast-milk exposure on human infant outcomes is extremely limited. The available data comes primarily from two sources:
Astley and Little (1990), Neurotoxicology and Teratology. This study of 136 breastfed infants found that marijuana exposure through breast milk in the first month of life was associated with a modest reduction in motor development at one year of age, as measured by the Bayley Scales of Infant Development. However, the study could not fully control for prenatal cannabis exposure, and the effect was small.
The Tennes et al. (1985) study. An older study that found no significant differences in developmental outcomes between infants exposed to cannabis through breast milk and unexposed infants at one year. However, this study was small and the cannabis available in the 1980s was far less potent than contemporary products.
Critical evidence gaps. No study has examined the effects of current high-potency cannabis exposure through breast milk on long-term neurodevelopmental outcomes using modern assessment tools. The animal data raises concerns that have not been adequately addressed in human research.
What Medical Organizations Recommend
Every major medical organization that has issued guidance on this topic recommends against cannabis use during breastfeeding.
| Organization | Recommendation |
|---|---|
| American Academy of Pediatrics (AAP) | Advises against cannabis use during breastfeeding |
| American College of Obstetricians and Gynecologists (ACOG) | Recommends discontinuation during pregnancy and lactation |
| Academy of Breastfeeding Medicine (ABM) | Advises against; supports continued breastfeeding if use occurs |
| Centers for Disease Control and Prevention (CDC) | Advises avoiding cannabis during breastfeeding |
| World Health Organization (WHO) | Advises against substance use during breastfeeding |
An important nuance: the Academy of Breastfeeding Medicine (ABM) explicitly notes that cannabis use should not be considered a contraindication to breastfeeding. Their 2015 protocol states that while mothers should be counseled against cannabis use, the benefits of breastfeeding generally outweigh the potential risks of low-level THC exposure through breast milk. This is a critical distinction — advising against cannabis use is not the same as advising against breastfeeding if cannabis has been used.
The Risk-Benefit Calculus
The honest assessment of the current evidence is this:
What we know:
- THC readily transfers into breast milk at concentrations exceeding plasma levels
- THC persists in breast milk for 4–6 days after a single use
- Infants do absorb THC from breast milk (confirmed by urine/stool detection)
- The endocannabinoid system is active and important in infant brain development
- Animal studies show that postnatal cannabinoid exposure produces neurodevelopmental changes
What we do not know:
- The actual neurodevelopmental effects of low-level THC exposure through breast milk in human infants
- Whether a dose threshold exists below which exposure is harmless
- Whether occasional use poses a meaningfully different risk than regular use
- The specific long-term outcomes (cognitive, behavioral, psychiatric) of exposed infants
What is clear:
- The precautionary principle strongly favors avoiding cannabis during breastfeeding
- The absence of proven harm is not the same as proof of safety
- Breastfeeding itself has substantial, well-documented health benefits that should not be abandoned lightly
Practical Guidance
For breastfeeding parents who use cannabis: The evidence-based recommendation is to abstain during the entire period of breastfeeding. If abstinence is not possible, reducing frequency and amount of use reduces total infant exposure. Edible and oral forms may result in lower peak breast milk THC concentrations than inhaled forms (due to lower peak blood levels from oral administration), but this has not been specifically studied in the context of lactation.
For healthcare providers: Approach cannabis-using breastfeeding parents without judgment. Shaming or threatening to involve child protective services drives parents away from honest disclosure and medical guidance. Provide clear information about what is known and unknown, support continued breastfeeding, and help parents make informed decisions.
For researchers: This is an area that desperately needs large-scale, prospective, longitudinal studies tracking developmental outcomes in infants exposed to quantified levels of THC through breast milk. The current evidence base is grossly insufficient for the public health significance of the question.
For everyone: The rapid increase in cannabis use among breastfeeding parents, combined with the dramatic increase in product potency, makes this a public health question of growing urgency. The answer to that question requires better data, not assumptions in either direction.
The current best advice remains straightforward: do not use cannabis while breastfeeding. The data supporting this advice is sound, even if incomplete. And for parents who do use cannabis while breastfeeding, the consensus is equally clear: continue breastfeeding, because its benefits are substantial and well-documented, and work with a healthcare provider to minimize risk.