A widely shared claim, based on the Jamaica study, says that cannabis is safe in pregnancy, and even good for the baby. It usually comes up as a question, “Is cannabis good for pregnant women?”, and the answer given relies mostly on the work of Dr Melanie Dreher, an American nurse-researcher who studied cannabis-using mothers in rural Jamaica. It is also usually paired with the argument that earlier North American studies showing harm had failed to separate cannabis from alcohol, tobacco and other drugs.
The study is Dreher, Nugent and Hudgins, “Prenatal Marijuana Exposure and Neonatal Outcomes in Jamaica: An Ethnographic Study”, published in Pediatrics in 1994. It compared 24 infants of regularly cannabis-using mothers with 20 infants of non-users. At three days old there were no significant differences. At one month, the exposed infants scored better on some neonatal assessments. The authors themselves linked that advantage mainly to the social and economic characteristics of the mothers who used cannabis, not to cannabis itself.
The criticism of the older literature was, and remains, a fair one. The conclusion drawn from it is not. A total sample of 44 infants in one rural setting cannot carry the weight the claim puts on it, and it does not survive contact with what has been published since.
What the current evidence base looks like
A systematic review and meta-analysis by Lo et al., published in JAMA Pediatrics in 2025, pooled 51 studies covering 21,146,938 participants. Pooled adjusted odds ratios:
| Outcome | Odds ratio (95% CI) | Studies | Certainty |
|---|---|---|---|
| Low birth weight (<2500 g) | 1.75 (1.41–2.18) | 20 | Moderate |
| Preterm birth (<37 weeks) | 1.52 (1.26–1.83) | 20 | Moderate |
| Small for gestational age | 1.57 (1.36–1.81) | 12 | Moderate |
| Perinatal mortality | 1.29 (1.07–1.55) | 6 | Low |
The certainty rating is the part to notice. This review updates the same team’s earlier one (Lo et al., Cannabis and Cannabinoid Research, 2024), which rated the evidence very low or low certainty. The 2025 update had more studies, consistent findings and signs of a dose–response pattern in heavy users. On that basis it raised low birth weight, preterm birth and small for gestational age to moderate. Perinatal mortality stayed at low. Over time, this literature has moved consistently towards more confidence in harm, not less. That is the opposite of what the Jamaica-based claim implies.
Tobacco: what the adjustment does and does not settle. Tobacco co-use is the classic confounder here. Cannabis users are more likely to smoke tobacco, and tobacco independently causes low birth weight and preterm birth. The 2025 review only included studies that adjusted for confounders, with tobacco use as the minimum. The authors report that the links with preterm birth, small for gestational age and low birth weight held after that adjustment. That does not close the question. An adjustment is only as good as the tobacco data behind it, and those data are usually self-reported, so some residual confounding is possible.
Other honest limitations of this literature: exposure is almost always self-reported, and dose is rarely measured. The review’s heavy-use analyses point towards a dose–response pattern, but there is no well-defined dose–response curve. These are real weaknesses. They are not a reason to conclude the opposite.
What the professional bodies say
The American College of Obstetricians and Gynecologists issued Clinical Consensus No. 10, “Cannabis Use During Pregnancy and Lactation,” in September/October 2025. Its recommendations, as reported in ACOG’s own release and in press coverage of it:
- Advises against cannabis use during pregnancy and lactation
- Recommends universal screening of pregnant patients, by validated questionnaire rather than by urine toxicology
We should be precise about our sourcing here. ACOG’s own page and the Obstetrics & Gynecology full text were behind access controls we could not get through. The recommendations above are checked against ACOG’s public news release and mainstream coverage of it. The primary citation is Obstet Gynecol, Clinical Consensus No. 10, 2025, and anyone advising patients should read the original rather than this summary.
The screening detail is worth flagging because it protects patients: ACOG recommends asking, not testing. Urine toxicology in pregnancy carries legal and child-protection consequences that fall unevenly across populations, and a professional body recommending against it is making a deliberate choice.
Breast milk: what is actually measured
The breast-milk data have improved a lot recently. They also undermine the advice most often given: “wait a few hours and pump and dump.”
The Lactation and Cannabis (LAC) study (Holdsworth et al., Breastfeeding Medicine, 2024) followed 20 breastfeeding participants who used cannabis at least weekly and had infants under six months old:
- Each participant gave a baseline milk sample after at least 12 hours without cannabis, then five samples over 8–12 hours after use. According to the university’s release, THC was found in every sample, including the baseline ones. In a regular user, milk is not THC-free even after a 12-hour break.
- After a single use, THC in milk peaked at a median of 120 minutes (9 participants). The timing varied widely: press coverage of the study gives roughly 30 minutes to 2.5 hours, and the researchers said there was no consistent time when the concentration peaked and started to fall.
- Repeated use led to build-up (bioaccumulation) in most participants: the more often cannabis was used during the day, the higher the overall THC exposure in milk.
- The researchers estimated that infants took in an average of about 0.07 mg of THC per day, and said it is not known whether this affects the developing infant.
The milk-to-blood comparison comes from older, very small studies, not from the LAC study. A 1982 report by Perez-Reyes and Wall found THC in milk at concentrations several times higher than in the mother’s plasma. The “about seven times higher” figure often quoted traces back to that small early literature, so treat it as a rough indication, not a precise ratio.
This is why “pump and dump” does not work. The strategy assumes a peak you can wait out. There is no reliable peak, the baseline is not zero in a regular user, and repeated use builds up.
The LactMed database adds data on how long THC lasts in milk: it can be detected from 6 days to more than 6 weeks after use. Modelled half-life estimates range from 12 to 39 hours, with one estimate averaging about 27 hours, and median modelled clearance is 2.5 to 8 days.
What is known about effects on the infant
LactMed’s summary of the infant literature is genuinely mixed and we will report it as such:
- Occasional maternal use: one study found “no discernable effects on breastfed infants”
- Daily or near-daily use: one study found it “might retard the breastfed infant’s motor development, but not growth or intellectual development”
- Case reports in heavily exposed infants describe seizures, somnolence, and apnoeic episodes
- A registry study suggested postnatal cannabis use may be linked to autism spectrum disorder risk, though that research did not confirm whether the infants were breastfed
Two more effects worth knowing:
- Cannabis affects the milk-producing hormone prolactin in inconsistent ways and may reduce milk supply and shorten breastfeeding. Studies after legalisation in Colorado found mothers breastfed for a shorter time.
- Smoking anything near an infant is separately harmful, whatever is being smoked. LactMed states plainly that no one should smoke cannabis near infants.
Why this research is so hard to do
It helps to understand why the evidence looks the way it does. It explains both why the optimism of the Jamaica-based claim was once possible and why it no longer holds.
You cannot randomise pregnant women to a substance suspected of causing harm. Better funding would not fix this; it is an ethical boundary. Every study in the meta-analysis above is observational, which means every one of them is open to confounding.
Exposure measurement is poor. People under-report their own use. Product strength has risen a lot over the decades these cohort studies cover, so “cannabis use” in a 1990 cohort and in a 2020 cohort are not the same exposure.
Accounts that circulate with the claim also say the Jamaica research lost its funding because of its null results. We could not verify that, and even if true it would not change what a 44-infant study can show.
The Thai context (as of October 2026)
Under the Notification of the Ministry of Public Health Re: Controlled Herbs (Cannabis) B.E. 2568, issued 26 June 2025, cannabis flower is a controlled herb. The rules:
- Flower may be dispensed only against a prescription on form ภ.ท. 33 (PT 33) from an authorised practitioner. Clause 4 lists seven categories: medical doctors, dentists, pharmacists, Thai traditional medicine practitioners, applied Thai traditional medicine practitioners, Chinese medicine practitioners and folk healers.
- Each prescription covers no more than 30 days and is single use (no refills).
- Flower that is sold or exported must come from a GACP-certified cultivation site.
- Online sale and advertising of flower are prohibited, and recreational use is not permitted.
On pregnancy and breastfeeding specifically: the B.E. 2565 notification that this one replaced explicitly banned selling controlled cannabis to pregnant and breastfeeding women, and to people under 20. The published summaries of the B.E. 2568 text that we checked do not repeat that list, and we could not confirm a clause in the 2568 notification that does. In practice, flower can reach a pregnant or breastfeeding woman legally only through a PT 33 prescription. Press reports on enforcement guidelines say that selling to pregnant or breastfeeding women without a PT 33 is grounds for revoking a shop’s licence. If you are pregnant or breastfeeding, the prescriber needs to know.
The rules are changing. As of October 2026, the Cannabis Control Act is still a draft: Cabinet approved it on 22 September 2026 and sent it to the House. According to press reports, the draft would add an explicit ban on sales to under-20s. Two Ministerial Regulations B.E. 2569 have also tightened licensing: one on cannabis/hemp extracts (in force 26 April 2026) and one on controlled-herb licences, No. 2 (in force 30 April 2026). They do not change the PT 33 prescription rules. Check the forum’s law topics (1330 and 136) for the current position before relying on this section.
Nausea and vomiting of pregnancy
This is the reason people most often give for using cannabis in pregnancy, and it deserves a direct answer.
Nausea and vomiting in early pregnancy is miserable and sometimes severe, and the urge to reach for something that works is not irrational. Three things are worth considering alongside it:
- There are effective, evaluated treatments for nausea and vomiting of pregnancy, including hyperemesis gravidarum, that a doctor can prescribe.
- Cannabinoid hyperemesis syndrome (CHS) causes repeated bouts of severe vomiting in regular cannabis users. In pregnancy it can be mistaken for ordinary pregnancy sickness or hyperemesis gravidarum, especially when cannabis use is not mentioned, and that delays the right treatment. The key treatment for CHS is stopping cannabis; symptoms go away when use stops (Flament, Scius and Thonon, International Journal of Emergency Medicine, 2020). If vomiting persists or gets worse with continued use, CHS needs to be considered.
- The first trimester, when nausea peaks, is also when the baby’s organs are forming.
We could not find a Thailand-specific clinical guideline for nausea and vomiting of pregnancy, so we are not going to invent one. In Thailand, the natural place to raise this is antenatal care (ANC), for example at a government hospital. Tell the ANC team about any cannabis use, including products bought as CBD, so they can consider CHS and treat the nausea properly.
Where this leaves us
The current position of the professional bodies is to advise against cannabis use in pregnancy and breastfeeding. The observational evidence behind that position is imperfect in the ways described above, and the associations in the meta-analysis are moderate rather than dramatic. Nobody should be told that a single exposure will harm their baby, because that is not what the data say either.
But the honest version is not the Jamaica-based version. That claim rested on a 44-infant study in one place at one time, and the more than twenty-one million participants studied since do not support it.
Anyone making this decision should make it with their own clinician, with the actual numbers in front of them, and without either the reassurance or the alarm that dominate this topic online.
Sources: Lo JO, Ayers CK, Yeddala S, Shaw B, Robalino S, Ward R, Kansagara D. “Prenatal Cannabis Use and Neonatal Outcomes: A Systematic Review and Meta-Analysis.” JAMA Pediatrics 2025;179(7):738–746. doi:10.1001/jamapediatrics.2025.0689 · Lo JO, Shaw B, Robalino S, et al. “Cannabis Use in Pregnancy and Neonatal Outcomes: A Systematic Review and Meta-Analysis.” Cannabis and Cannabinoid Research 2024;9(2):470–485. doi:10.1089/can.2022.0262 · Dreher MC, Nugent K, Hudgins R. “Prenatal Marijuana Exposure and Neonatal Outcomes in Jamaica: An Ethnographic Study.” Pediatrics 1994;93(2):254–260. doi:10.1542/peds.93.2.254 · ACOG Clinical Consensus No. 10, “Cannabis Use During Pregnancy and Lactation”, Obstetrics & Gynecology, 2025 (accessed via ACOG news release) · Holdsworth EA, Berim A, Gang DR, Williams JE, Smith CB, Caffé B, Brooks O, Barbosa-Leiker C, McGuire MA, McGuire MK, Meehan CL. “Human Milk Cannabinoid Concentrations and Associations with Maternal Factors: The Lactation and Cannabis (LAC) Study.” Breastfeeding Medicine 2024;19(7):515–524. doi:10.1089/bfm.2024.0021 · Perez-Reyes M, Wall ME. “Presence of Δ9-tetrahydrocannabinol in human milk.” N Engl J Med 1982;307:819–820 · Flament J, Scius N, Thonon H. “Cannabinoid hyperemesis syndrome in the pregnant patient: clinical case and literature review.” International Journal of Emergency Medicine 2020;13:52. doi:10.1186/s12245-020-00311-y · “Cannabis”, Drugs and Lactation Database (LactMed), NCBI Bookshelf · Notification of the Ministry of Public Health Re: Controlled Herbs (Cannabis) B.E. 2568.