Every other page in the Women's Health cluster asks which cannabinoid. This one asks whether at all, and for anyone feeding a baby their own milk the answer from the American College of Obstetricians and Gynecologists, the American Academy of Pediatrics, and the NIH's drug-and-lactation database is no. The reason is chemistry. THC dissolves in fat, milk is a fat delivery system, and studies have found THC in milk anywhere from six days to more than six weeks after the last use, depending on how heavily the person had used before stopping.
That range is the spine of this page. Every workaround the market offers, from pump and dump to "clean" hemp CBD, runs into it. The position recommends nothing: if you are breastfeeding, edibles are off the table for as long as you are. If you are not, the milk problem disappears and a different one stays. The page exists because this population searches, and what it finds is mostly written by people selling something.
Is it safe to eat THC edibles while breastfeeding?
No, by the standard of every professional body that has looked. ACOG Committee Opinion 722 discourages marijuana use during lactation because the data to judge infant effects do not exist. The AAP clinical report, reaffirmed in April 2026, strongly encourages complete abstinence while breastfeeding. LactMed summarizes the guidelines the same way.
The absence of data is the finding, and it will stay that way, since no ethics board will approve a trial that doses a nursing infant with THC. What exists is one old study. Astley and Little followed 136 breastfed infants to one year, half exposed to cannabis through milk and half matched on maternal alcohol and tobacco use, and found exposure in the first month associated with lower motor development scores at twelve months. That was 1990, with far weaker flower and mothers who also drank and smoked, so it proves less than it sounds like. It is also the study the AAP still cites.
The market has not waited: products with postpartum and mama on the label are on shelves, and the advice at the counter has been measured. In 2018 Dickson and colleagues had a caller phone 400 randomly selected Colorado dispensaries, say she was eight weeks pregnant with morning sickness, and ask what to take. Sixty-nine percent, 277 of 400, recommended a cannabis product. Only 31.8% suggested a doctor without being prompted. That is pregnancy rather than lactation, at the same counter. The budtender is a retail employee. The question was medical.
How long does THC stay in breast milk?
Six days to more than six weeks, depending on how heavily the person used before stopping. LactMed gives that range. The short end comes from Bertrand 2018, where THC was measurable about six days after last use. The long end comes from Wymore 2021, where prenatal users who quit at delivery still had THC in milk at six weeks.
Bertrand and colleagues drew on Mommy's Milk, a human milk research repository, and analyzed 54 samples from 50 breastfeeding women who reported cannabis use. THC was detectable in 34 of the 54 (63%), at a median of 9.47 ng/mL, up to roughly six days after last reported use. LactMed puts the milk half-life in that dataset at about 27 hours. That is the short end, and it describes mostly occasional users.
Wymore and colleagues enrolled 25 women who had used cannabis more than twice a week during pregnancy, mostly by smoking, who planned to breastfeed and abstain for six weeks. Twelve were confirmed abstinent by plasma testing, across 402 matched plasma and milk samples, and every one of them had detectable THC in milk throughout. Among the seven who stayed abstinent past five weeks, the milk half-life averaged 17 days, and the modelled time to elimination ran past the six-week endpoint. Median milk THC was 3.2 ng/mL in week one, 5.5 in week two, and 1.9 in week six. The milk got stronger in the second week of not using.
Baker and colleagues had eight exclusively breastfeeding women abstain for 24 hours, then smoke 100 mg of a preweighed, standardized flower at 23.18% THC from one preselected dispensary. The estimated relative infant dose was 2.5% of the maternal dose, ranging from 0.4% to 8.7%. Moss and colleagues sampled 20 nursing mothers at two weeks and two months postpartum and found a median milk THC of 27.5 ng/mL against 3.7 in plasma, a milk-to-plasma ratio of 7. Milk did not dilute the drug. It concentrated it.
Almost all of this is inhaled cannabis. Baker's participants smoked by design, Bertrand's and Wymore's mostly smoked, and Moss had three edible users whose milk looked like the smokers'. Nothing about the oral route makes milk safer. If anything the pharmacology points the other way: an edible produces a plasma curve that peaks lower and runs several hours longer than a smoked dose, which means more hours during which milk is drawing THC from the blood. That last sentence is an inference from the plasma curve rather than a milk measurement, and it is labelled as one, because nobody has run the edible-specific milk study.
Can you pump and dump after an edible?
No. Pumping and discarding works for alcohol because alcohol leaves milk on the same clock it leaves blood, a matter of hours. THC is stored in body fat, and milk is refilled from that store, so milk levels track fat over days to weeks. Emptying the breast removes one batch; the next batch draws from the same reservoir.
The Wymore numbers are the cleanest demonstration. Those women pumped two to five times a week for six weeks with no new cannabis going in, and the milk concentration rose between week one and week two anyway. LactMed notes that milk THC content correlates with milk fat content. Pump and dump is folklore that borrowed alcohol's pharmacology and applied it to a molecule with the opposite behaviour. It empties the breast. It does not empty the parent.
Is CBD safe while breastfeeding?
Nobody knows, and the FDA advises against it. No lactation study of CBD on its own exists. LactMed's cannabidiol record says it has not been studied in nursing women but has been measured in the milk of cannabis users. In Moss 2021, 13 of 20 mothers had CBD in their milk and none had reported using a CBD product.
The FDA's consumer update covers CBD by name. It says there is no comprehensive research on CBD's effects on a breastfed baby, that the agency expects some CBD to transfer through milk, and that CBD products can be contaminated with THC. Full-spectrum hemp products contain THC by design, and Bertrand's repository found CBD in five samples, sitting alongside the THC. The one CBD drug the FDA has approved carries warnings for liver toxicity and sedation, a strange list to hand to someone whose body is currently the food supply.
Do edibles help with postpartum sleep, pain, or anxiety?
There is no trial of any cannabis product for postpartum sleep, pain, mood, or anxiety. Every claim on a product page for this population is borrowed from other populations. And the one use with real evidence elsewhere, sleep, collides with the AAP's safe-sleep rules, because a sedated parent and a newborn are the pairing those rules exist to prevent.
The 2022 AAP safe-sleep statement names marijuana in its list of substances to avoid during pregnancy and after birth. It describes bed sharing as especially hazardous when the parent's alertness is impaired by sedating medication or substances, and it puts couches and armchairs at 22 to 67 times the baseline risk of infant death. A standard edible runs six to eight hours. A newborn feeds every two to three. The AAP wrote its bed-sharing section for the adult holding a six-week-old at 3am, and that adult is the reader here.
What changes if you are not breastfeeding?
The milk problem disappears and the safe-sleep problem stays. A parent who is formula feeding, or who has weaned, is pharmacologically an ordinary adult, and the closest profile on this site is the perimenopause anxiety page: CBD-forward, THC held low, with the caveat that THC calms at small doses and does the opposite at larger ones.
That biphasic caveat lands harder on a brain running on four hours of broken sleep than on the 47-year-old it was written for. Sleep deprivation narrows the gap between the dose that settles you and the dose that hands you a racing heart, and nobody has measured where that gap sits in the first postpartum year. The same six to eight hour window applies, the night feeds still fall inside it, and the person taking the 2am shift should not be the person who took the gummy. The dosing guide covers starting low for anyone whose tolerance was last calibrated years ago, which after a pregnancy is nearly everyone.
What every postpartum reader should know before any edible
Never bed-share, and never feed on a couch or chair, after an edible. The AAP's 2022 recommendations treat a parent with impaired alertness as a bed-sharing hazard. A six-to-eight hour window overlaps every night feed. Back to the crib or bassinet after feeding, every time.
Edibles are not a treatment for postpartum depression or anxiety. The screen is the Edinburgh Postnatal Depression Scale, and the treatments are talk therapy and medication. Thoughts of harming yourself or the baby are an emergency. The National Maternal Mental Health Hotline is 1-833-TLC-MAMA (1-833-852-6262), free, confidential, 24 hours a day, in English and Spanish, by call, text, or chat.
Do not drive with an infant after an edible. The impairment lasts as long as the window does, and a car seat does not change that arithmetic.
Store edibles as if a toddler will find them, because half the time one does. The 25-year poison center study on our child ingestions page found edibles behind nearly half of all recreational drug ingestions by children under six. Original child-resistant container, locked, out of sight and reach.
Tell whoever prescribes for you. No specific interaction between cannabinoids and the drugs common in the first postpartum year is documented well enough to name here. Anyone on an SSRI or a post-cesarean opioid should have that conversation before the first piece.
Urine tests say THC lingers. Milk says where it goes. A breastfeeding parent is the one person whose "it's out of my system by now" can be checked against a fluid someone else is about to drink, and in the only cohort that checked weekly for six weeks, it never was. The rest of the cannabis market gets to round that number down. This reader does not.