Cannabis Use in Pregnancy – Context, Evidence, and Clinical Reality
Authored by Felecia Dawson, MD, ABOG and Genester Wilson-King, MD, FACOG
Edited by Jahan Marcu, PhD
In October 2025, the American College of Obstetricians and Gynecologists updated its guidance on cannabis use during pregnancy and lactation. The revision deserves some credit. Its tone is more patient-centered. It acknowledges inequities. It moves, at least in part, away from punishment and toward conversation.
But tone is not the same thing as clinical usefulness. A compassionate document can still be scientifically incomplete. A nonjudgmental recommendation can still leave clinicians with little more than “tell patients to stop” and “follow your state reporting laws.” For patients living in the real world, and for clinicians trying to care for them, that is not enough.
The Society of Cannabis Clinicians offers this response to place the guidance in historical, scientific, and clinical context, and to call for a more accurate, humane, and practical approach to cannabis use during pregnancy and lactation.
Cannabis Did Not Appear Yesterday
One limitation of modern cannabis guidance is that it often behaves as though cannabis began with legalization, dispensaries, or adolescent risk surveys. It did not.
Cannabis has been used as a botanical medicine across cultures for thousands of years, including in contexts related to women’s health, childbirth, nausea, pain, bleeding, and reproductive disorders. Historical medical literature describes oral, inhaled, topical, vaginal, rectal, and other preparations used for conditions ranging from dysmenorrhea and difficult labor to nausea of pregnancy and postpartum hemorrhage.
Cannabis also had a place in Western medicine. It entered American medical practice in the nineteenth century and was listed in the U.S. Pharmacopeia beginning in 1850. Its disappearance from medicine was not the result of a clean scientific verdict. It was the result of politics, racialized drug policy, professional disruption, and later Schedule I classification.
That history does not prove cannabis is safe in pregnancy. History is not a clinical trial. But it does remind us that the current evidence base did not emerge from a neutral scientific landscape. It emerged from a century of prohibition, stigma, and research restrictions.
The Evidence Base Is Not as Clean as the Guidance Implies
The central problem with prenatal cannabis research is not that it finds risk. Risk deserves serious attention. The problem is that the evidence is often treated as cleaner, more specific, and more actionable than it actually is.
Many studies rely on self-report in highly stigmatized settings. Many lack reliable dose, timing, frequency, route-of-administration, or product-composition data. Many struggle to separate cannabis exposure from tobacco, alcohol, other drugs, poverty, stress, nutrition, trauma, limited prenatal care, and surveillance bias. Some preclinical studies use synthetic cannabinoids that are pharmacologically quite different from cannabis products people actually use.
This does not make the literature meaningless. It means clinicians should interpret it with the same discipline they would bring to any other complicated exposure: What was used? How much? When? By whom? In what form? With what other exposures? Under what social and clinical conditions?
A blanket statement that cannabis has “no medical indication” during pregnancy may be administratively tidy, but it is clinically incomplete. Patients report using cannabis for nausea, vomiting, pain, sleep, anxiety, appetite, and, in some cases, severe conditions such as hyperemesis gravidarum. Whether cannabis should be recommended is a separate question. But pretending these therapeutic motives do not exist is not patient-centered medicine. It is clinical avoidance with better stationery.
Omitted Evidence Matters
The current guidance also gives insufficient attention to studies that complicate the dominant narrative.
Some higher-quality reviews and cohort studies, after controlling for relevant confounders, have not found clear evidence that prenatal cannabis exposure alone produces clinically meaningful long-term cognitive or neurodevelopmental impairment. These findings should not be oversold. They do not prove safety. But they do challenge the habit of treating every association as though it were a settled causal fact.
Evidence-based medicine requires more than collecting the studies that confirm our anxieties. It requires looking directly at uncertainty and saying, with some humility, what is known, what is plausible, what is unknown, and what may have been exaggerated by bias, poor exposure assessment, or moral panic.
The Endocannabinoid System Is Not Optional Background
Perhaps the most striking omission is the limited treatment of the endocannabinoid system.
The ECS is not a cannabis trivia fact. It is a major biological signaling system involved in implantation, placentation, immune modulation, nausea, pain, mood, appetite, stress regulation, and neurodevelopment. It is the system through which cannabinoids exert many of their effects.
A guidance document on cannabis in pregnancy that barely discusses the ECS is like a cardiology document that mentions the heart only in a footnote.
Clinicians cannot meaningfully evaluate risk, benefit, timing, mechanism, or patient-reported effects without some working understanding of the ECS. Yet medical education continues to underprepare clinicians in this area. The result is predictable: patients ask practical questions, and clinicians are left with slogans.
Route, Dose, and Product Type Matter
The National Academies’ 2017 review found substantial evidence linking maternal cannabis smoking with lower birth weight, limited evidence for some pregnancy complications and NICU admission, and insufficient evidence to support or refute many long-term cognitive, academic, or behavioral outcomes.
The word “smoking” matters.
Combustion is not the same thing as vaporization. A regulated oral preparation is not the same thing as a gas-station hemp product. A full-spectrum botanical extract is not the same thing as a synthetic cannabinoid. THC is not CBD. CBD is not CBG. A contaminated vape cartridge is not a standardized capsule. Dose matters. Timing matters. Frequency matters. Product chemistry matters.
Medicine has learned this lesson before. Hormone therapy was once flattened into a single risk category until later evidence clarified that formulation, dose, route, and patient characteristics radically altered the risk-benefit profile. Cannabis science is approaching a similar inflection point. If we continue treating all exposures as interchangeable, we will continue producing guidance that is simultaneously overconfident and under-informative.
Harm Reduction Cannot Be an Afterthought
The updated guidance gestures toward motivational interviewing and supportive care. That is progress. But the harm-reduction framework remains underdeveloped.
For some patients, cessation will be realistic and appropriate. For others, it will not happen immediately, or at all. Some patients use cannabis after other therapies fail, are unaffordable, are inaccessible, or produce intolerable side effects. Some use cannabis instead of substances with better-known and more severe risks. Some live in states where disclosure may trigger surveillance, reporting, or family separation.
A clinically useful document must help providers care for those patients too.
At minimum, clinicians need practical guidance on reducing risk when cannabis use continues: avoiding combustion, avoiding high-potency products, avoiding synthetic cannabinoids, avoiding unregulated smoke-shop and vape-shop products, understanding contamination risks, considering third-party testing, assessing polysubstance use, and documenting counseling in a way that protects rather than punishes patients.
A policy that only works for the most compliant patient in the safest legal environment is not a policy. It is a wish.
Equity Requires More Than Acknowledgment
The updated guidance rightly acknowledges racialized surveillance and unequal enforcement. But acknowledgment without operational guidance is thin protection.
Clinicians need concrete recommendations on how to avoid turning prenatal care into a pipeline for punishment. This includes careful consideration of when testing is clinically necessary, how consent is obtained, how results are interpreted, how mandatory reporting laws are explained, and how patients are protected from unnecessary legal and social harm.
Patients should not have to choose between honest conversation and self-preservation.
Progress, But Not Enough
The SCC recognizes meaningful improvements in the updated guidance. The shift toward nonpunitive language matters. Support for continued breastfeeding in some circumstances matters. Recognition of inequity matters. Encouraging clinicians to initiate conversations matters. Using the term cannabis rather than marijuana matters.
These are real steps forward.
But the next step must be better science, better clinical nuance, and better harm reduction.
Clinicians deserve guidance that helps them distinguish between cannabis products, routes of administration, dose patterns, and patient circumstances. Patients deserve counseling that is honest without being punitive, cautious without being simplistic, and grounded in biology rather than inherited stigma.
Cannabis use during pregnancy is not a topic for slogans. It is a topic for humility, careful evidence review, clinical judgment, and compassion.
ACOG has moved in the right direction. Now the medical community must go further.
Women deserve better. Pregnant patients deserve better. Clinicians deserve better. Evidence-based medicine demands better.
May 18, 2026