- THC in milk is 8× blood concentration: THC’s high lipophilicity causes it to concentrate in milk fat at 8-fold higher levels than maternal blood plasma.
- Detectable for 6 days: A single cannabis use episode results in detectable THC in breast milk for up to 6 days; chronic daily use creates persistent elevated baseline levels.
- Infant dose estimate: Infants receive approximately 2.5% of the maternal weight-adjusted dose per feeding — a meaningful pharmacological dose for a developing brain.
- Pump-and-dump does not work: Unlike alcohol, THC cannot be cleared from milk by pumping and discarding, because THC is stored in breast tissue fat and continuously released into new milk production.
- ECS-dependent brain development: The neonatal period is characterised by intense endocannabinoid-dependent synaptogenesis and pruning; exogenous THC disrupts these processes.
- Longitudinal data: Children with prenatal/early postnatal cannabis exposure show impaired executive function and attention at ages 3–10 in multiple cohort studies.
- Universal prohibition: CDC, AAP, ACOG, and the Academy of Breastfeeding Medicine all advise against cannabis use during breastfeeding with no safe threshold defined.
Why THC Concentrates in Breast Milk
THC (delta-9-tetrahydrocannabinol) is a highly lipophilic compound with a log P value of approximately 7.2 — meaning it partitions strongly into fat-containing tissues relative to aqueous (water-based) environments. Blood plasma is primarily aqueous; breast milk, particularly mature hindmilk, contains 3–5% fat by volume. This fat content creates a pharmacokinetic reservoir that actively concentrates lipophilic substances from maternal circulation.
The milk:plasma ratio for THC has been measured at approximately 8:1 in multiple analytical studies, including Baker et al. (2018) which used highly sensitive liquid chromatography-mass spectrometry (LC-MS/MS) to quantify cannabinoids in matched maternal plasma and milk samples. This means that at any given maternal blood THC concentration, milk contains 8 times more THC per unit volume.
Furthermore, THC accumulates in breast tissue adipocytes (fat cells) over time with repeated use. This tissue reservoir slowly releases THC back into milk production regardless of whether the mother actively uses cannabis at that moment. This pharmacokinetic property is critical to understand: casual or intermittent cannabis users will have milk containing measurable THC for days after each use; daily users maintain persistently elevated milk THC levels.
Research Data: THC Concentrations in Human Breast Milk
| Study | N (samples) | THC Detection Rate | Median THC in Milk | Range | Duration Detectable |
|---|---|---|---|---|---|
| Baker et al. (2018), Obstetrics & Gynecology | 54 samples, 25 mothers | 63% | 9.1 ng/mL | 0–323 ng/mL | Up to 6 days post-use |
| Moss et al. (2021), JAMA Pediatrics | 40 samples, 20 mothers | 75% | 14.5 ng/mL | 0.2–289 ng/mL | Up to 6 days post-use |
| Camara-Lemarroy et al. (2021) | Review of 11 studies | Variable (42–100%) | Variable | 2–8 days |
The wide concentration range reflects differences in maternal use frequency, recency of use at sample collection, and milk collection timing (foremilk vs hindmilk — hindmilk has higher fat and thus higher THC). The estimated infant dose of 2.5% maternal weight-adjusted dose per feeding (from Perez-Reyes & Wall, 1982, the reference basis for dose estimation) translates to approximately 0.1–8 ug/kg/day depending on maternal use pattern — a pharmacologically meaningful range for a developing central nervous system.
THCCOOH (THC carboxylic acid, the primary inactive metabolite detectable in urine drug screens) is also transferred into milk and has been detected at concentrations approximately 2–3× higher than THC itself, as it is less lipophilic and distributes more broadly.
The Developing Endocannabinoid System and Vulnerability to THC
The neonatal and infant period is characterised by extraordinary brain development rates: at birth, the human brain is approximately 25% of adult volume; by age 2, it reaches 80%. This period involves rapid synaptogenesis (formation of new synapses), synaptic pruning (elimination of redundant connections), myelination, and the formation of long-range cortical connectivity.
Endocannabinoid signalling is critically involved in all of these processes. CB1 receptors are expressed in the developing brain at higher densities than in the adult brain, and the endocannabinoid system serves as a key regulator of axonal guidance, synapse maturation, and neural circuit formation. Endogenous 2-AG, for example, directs pyramidal neuron axon orientation during cortical development through a process that requires precise spatiotemporal CB1 receptor activation.
Exogenous THC delivered through breast milk disrupts this precisely regulated system by: (1) providing non-degradable, persistent CB1 activation that overrides the on-demand endocannabinoid signalling governing development; (2) causing premature CB1 receptor internalisation in developing neurons; and (3) potentially altering the expression patterns of endocannabinoid metabolic enzymes during critical developmental windows that cannot be compensated for later.
Longitudinal Outcomes: What the Research Shows
Demonstrating specific harm from breastfeeding-period cannabis exposure is methodologically challenging because it requires separating in-utero exposure from postnatal breast milk exposure, and controlling for confounding socioeconomic and environmental variables. Most available data come from studies of prenatal exposure combined with early postnatal exposure:
| Study | Population | Outcome Measured | Finding | Child Age at Assessment |
|---|---|---|---|---|
| Day et al. (2006), Neurotoxicology and Teratology | 648 prenatal-exposed children (PHLAME cohort) | Executive function, memory | Impaired executive function; increased impulsivity | Age 10 |
| Goldschmidt et al. (2004) | 580 children (Pittsburgh cohort) | Delinquency, attention | Increased delinquent behaviour; attention deficits | Age 10–14 |
| Fried et al. (2003) | 70 prenatal-exposed (Ottawa cohort) | Executive function (Stroop, WCST) | Reduced cognitive flexibility; increased perseveration | Age 13–16 |
| Gunn et al. (2016, systematic review) | 24 studies reviewed | Multiple neurological outcomes | Consistent association: attention, executive function, anxiety in exposed children | Various (3–16) |
While these studies predominantly measure prenatal rather than pure postnatal (breastfeeding-only) exposure, the developmental neurobiology evidence establishes a clear mechanistic basis for harm through any route of infant THC exposure during the first months of life. The developing brain at 0–6 months of age is undergoing the same endocannabinoid-dependent processes that were ongoing in utero.
Official Medical Guidelines
| Organisation | Position on Cannabis During Breastfeeding | Key Statement |
|---|---|---|
| CDC (Centers for Disease Control) | Advises against; no safe amount | “THC can be passed to your baby through breast milk. Breast milk can contain THC for up to 6 days after use.” |
| AAP (American Academy of Pediatrics) | Advises against; formal policy statement | “[The AAP] strongly advises against use of marijuana during pregnancy and while breastfeeding.” (AAP Policy, 2018) |
| ACOG (American College of OB/GYN) | Advises against; counselling required | “Women who are pregnant or contemplating pregnancy should be encouraged to discontinue marijuana use.” (Committee Opinion 722) |
| Academy of Breastfeeding Medicine | Advises against; harm reduction counselling recommended | Position: no known safe exposure; where cessation is impossible, harm reduction and formula supplementation preferred over cessation of breastfeeding entirely |
Why Pump-and-Dump Does Not Work
A common misconception — extrapolated from the strategy used for alcohol during breastfeeding — is that pumping and discarding breast milk after cannabis use will eliminate infant exposure. This is pharmacologically incorrect for THC.
Alcohol is water-soluble, reaches equilibrium between blood and milk rapidly, and clears from both simultaneously. Once blood alcohol falls below a threshold, milk alcohol is also negligible. The pump-and-dump strategy works for alcohol because of this rapid, concentration-following distribution behaviour.
THC behaves completely differently: it is stored in breast tissue fat, not in the milk aqueous phase. New milk produced after the pump-and-dump will contain THC released from the fat reservoir in breast tissue. There is no timing-based strategy that eliminates THC from breast milk because the source is not current blood levels but accumulated adipose tissue stores. Baker et al. (2018) confirmed that THC remained detectable in milk across all collection time points for 6 days regardless of pump timing relative to cannabis use.
Safer Alternatives and Harm Reduction
For mothers who use cannabis for legitimate medical or psychological reasons (chronic pain, anxiety, insomnia, nausea) during the postpartum period, the following hierarchy of options is recommended in consultation with a healthcare provider:
- Full cessation during breastfeeding period is the only approach that eliminates infant THC exposure through milk. Medical alternatives for the underlying condition should be actively evaluated.
- Formula feeding or donor milk eliminates the infant exposure pathway while allowing continued maternal cannabis use if medically necessary.
- Mixed feeding (partial formula supplement) with maximally spaced intervals between cannabis use and breastfeeding may reduce (but not eliminate) exposure — only for mothers for whom full cessation or formula feeding is not possible (Academy of Breastfeeding Medicine harm reduction framework).
- CBD-only products (no THC): Available data on CBD in breast milk are limited; however, CBD’s distinct pharmacology (no psychoactivity, different metabolism) places it in a different risk category. The AAP notes insufficient data on CBD specifically but advises avoidance until research matures.
Mothers experiencing anxiety, pain, or insomnia postpartum should be offered non-cannabis pharmacological and non-pharmacological alternatives including: SSRIs/SNRIs (compatible with breastfeeding at most doses), cognitive behavioural therapy, physiotherapy, lactation support for pain, and sleep hygiene interventions.
Frequently Asked Questions
Yes. THC is highly fat-soluble and concentrates in breast milk at 8× maternal blood plasma levels. Baker et al. (2018) confirmed detectable THC in human milk for up to 6 days after last use. Infants receive approximately 2.5% of the maternal dose per feeding.
The CDC advises that breastfeeding mothers should not use cannabis, stating there is no known safe amount of infant THC exposure through breast milk. This position is shared by the AAP, ACOG, and Academy of Breastfeeding Medicine.
Up to 6 days after a single use episode, and longer with chronic daily use. Because THC accumulates in breast tissue fat and releases continuously into new milk production, pump-and-dump is not effective as a clearance strategy.
The developing endocannabinoid system regulates synaptogenesis and neural circuit formation. THC exposure disrupts CB1-mediated development. Longitudinal cohort studies link early cannabis exposure to impaired executive function, reduced attention, and increased anxiety risk in children assessed at ages 3–10.
Cannabis Use Prevalence in Breastfeeding Populations
Understanding the scope of cannabis use during the breastfeeding period is critical to designing effective public health communication. Prevalence data from the United States show increasing use rates coinciding with cannabis legalisation:
| Data Source | Period | Cannabis Use Prevalence (Postpartum) | Trend |
|---|---|---|---|
| NSDUH (National Survey on Drug Use and Health) | Various annual waves | 4.3% past-month use in past-year postpartum women (2018–2019) | Increasing year-over-year |
| PRAMS (Pregnancy Risk Assessment Monitoring System, CDC) | 2016–2019 | 5–10% self-reported use during pregnancy across monitored states | Increasing in legal states |
| California Preterm Birth Initiative | 2018 | 7.1% positive urine toxicology for THC at delivery | Higher than self-reported |
| Baker et al. (2018) cohort | Clinical sample | 63% of self-reported users had detectable THC in milk | Consistent across use patterns |
Toxicological testing consistently reveals substantially higher rates of cannabis exposure than self-reported surveys capture, reflecting social desirability bias and the stigma associated with disclosing cannabis use during pregnancy and breastfeeding. Healthcare providers should screen for cannabis use non-judgementally, using validated tools, to provide appropriate counselling.
Barriers to Cessation
The most commonly cited reasons for continued cannabis use during breastfeeding include: nausea management (particularly where pregnancy nausea extends postpartum), anxiety and depression management, insomnia, and the widely available but inaccurate online information suggesting that cannabis is safe during breastfeeding. Many users cite dispensary staff as a primary information source — staff who may not have medical training and who operate in a commercial context that creates perverse incentives toward minimising safety concerns.
Effective harm reduction requires addressing these underlying reasons with evidence-based alternatives, rather than simply communicating a prohibition message that does not address the underlying symptom driving use.
CBD-Specific Considerations During Breastfeeding
Many mothers who use cannabis during breastfeeding specifically use CBD products (oils, gummies, topicals) under the belief that CBD is entirely safe because it does not produce psychoactive effects. The available evidence on CBD during breastfeeding is sparse but warrants careful consideration:
The FDA specifically warned in 2019 and reiterated subsequently that CBD during pregnancy and breastfeeding has not been adequately studied to establish safety. The concern is not primarily psychoactivity (CBD produces none) but rather unknown effects on the developing infant’s endocannabinoid system, which CBD modulates through multiple non-CB1 receptor mechanisms including TRPV1, 5-HT1A, and GPR55.
Data on CBD concentrations in human breast milk are very limited. CBD is less lipophilic than THC (log P approximately 6.3 vs 7.2) but is still substantially fat-soluble and would be expected to partition into breast milk to a meaningful degree. The absence of human data does not indicate safety; it reflects the difficulty of studying pharmaceutical interventions in lactating populations.
The Academy of Breastfeeding Medicine’s clinical protocol on cannabis (ABM Protocol 21, 2021) specifically addresses CBD: “Until more safety data are available, mothers who are breastfeeding should avoid CBD products, particularly those of unknown purity or concentration.” If CBD is being used for a specific medical indication (anxiety, pain), safer alternatives with established lactation safety profiles should be explored with the patient’s healthcare provider.
Topical CBD products applied to areas away from the breast present a different risk profile: systemic absorption from topical application is minimal. While topical CBD should still be disclosed to a healthcare provider, the risk of meaningful infant breast milk exposure from topical-only use is substantially lower than from oral or inhaled CBD products.
Related guides: Cannabis and Pregnancy • The Endocannabinoid System • Cannabis and Heart Rate • Cannabis for Depression