Cannabis use during pregnancy is one of the most emotionally charged and scientifically complicated topics in maternal health. With legalization expanding across the United States and cannabis becoming increasingly normalized, more pregnant women are asking a straightforward question: what does the evidence actually say?

The answer is more nuanced than either side of the debate typically admits. Anti-cannabis advocates often present worst-case scenarios as settled science, while cannabis enthusiasts sometimes dismiss legitimate concerns as prohibitionist fearmongering. The truth, as with most things in pharmacology, lives somewhere in the messy middle.

Here’s what we know heading into 2026, based on a review of more than 50 peer-reviewed studies, meta-analyses, and systematic reviews.

The Scope of the Issue

According to the National Survey on Drug Use and Health (NSDUH), approximately 7% of pregnant women in the United States reported cannabis use in the past month as of the most recent data. That number has been climbing steadily since 2002, when it sat around 3.4%. Among women aged 18-25, the rate is higher still, with some state-level surveys placing it above 12%.

The reasons vary. Morning sickness tops the list. A 2022 study published in JAMA Network Open by Volkow et al. found that pregnant women with severe nausea were 3.7 times more likely to use cannabis than those without. Other commonly cited reasons include anxiety, insomnia, chronic pain, and appetite stimulation, all symptoms that cannabis is known to address in non-pregnant populations.

The challenge for researchers: most of these women are also dealing with other confounding factors, including socioeconomic stress, tobacco use, alcohol consumption, inadequate prenatal care, and mental health conditions. Isolating the specific effect of cannabis from this tangle of variables is extraordinarily difficult.

How THC Crosses the Placenta

THC (delta-9-tetrahydrocannabinol) is highly lipophilic, meaning it readily dissolves in fat. The placenta, which functions as the interface between maternal and fetal blood supply, does not act as a complete barrier to lipophilic compounds. Studies using ex vivo human placental perfusion models have demonstrated that THC crosses the placental barrier, though the transfer rate appears to be partial rather than complete.

A landmark 2020 study by Bara et al. published in Biological Psychiatry found that THC reaches the fetal brain, where it can interact with the developing endocannabinoid system. This is significant because the endocannabinoid system (ECS) plays a critical role in fetal neurodevelopment. CB1 receptors begin appearing in the fetal brain as early as 14 weeks of gestation and are involved in neuronal proliferation, migration, and synapse formation.

The concentration of THC reaching the fetus appears to be lower than maternal blood levels, but the developing fetal brain is considerably more sensitive to disruption than the adult brain. Even low concentrations of an exogenous cannabinoid could theoretically alter the trajectory of ECS-mediated development.

CBD also crosses the placenta, though its pharmacological profile is different from THC. Far less research exists on prenatal CBD exposure specifically, and no long-term developmental studies in humans have been completed.

What the Studies Show: Outcome by Outcome

The evidence varies dramatically depending on which outcome you examine. Some findings are relatively consistent across multiple large studies. Others are contradictory, poorly controlled, or based on animal models that may not translate to humans.

Explore our Pregnancy Research Evidence Review below to see the strength of evidence for each outcome.

Low Birth Weight

This is one of the more studied outcomes. A 2016 meta-analysis by Gunn et al. in BMJ Open, which pooled data from 24 studies, found that cannabis use during pregnancy was associated with a 77% increased risk of low birth weight (defined as under 2,500 grams). The adjusted odds ratio was 1.77 (95% CI: 1.04-3.01) after controlling for tobacco use.

However, a 2020 Canadian study by Luke et al. published in CMAJ that tracked over 660,000 pregnancies found that after adjusting for tobacco, alcohol, and sociodemographic factors, the association between cannabis and low birth weight weakened substantially. The study concluded that residual confounding, particularly from concurrent tobacco use, likely inflates the risk estimates in many earlier studies.

The honest assessment: there is moderate evidence suggesting cannabis may contribute to lower birth weight, but the magnitude of the effect is debated, and separating it from tobacco exposure remains a persistent methodological challenge.

Preterm Birth

The evidence here is similarly mixed. The same Gunn et al. meta-analysis found an association between cannabis use and preterm birth (before 37 weeks), with a pooled odds ratio of 1.29. But a 2021 systematic review by Marchand et al. in Addiction noted that when studies controlled adequately for poly-substance use and socioeconomic status, the association became statistically insignificant in several analyses.

A notable 2023 study from Kaiser Permanente by Young-Wolff et al. tracked over 340,000 pregnancies in California and found a modest but statistically significant association between prenatal cannabis use and spontaneous preterm birth, even after adjusting for tobacco and other covariates. This is one of the stronger pieces of evidence in the preterm birth literature because of the enormous sample size and access to detailed medical records.

Cognitive and Behavioral Development

This is where the research gets particularly complicated, and where the stakes feel highest. Two longitudinal cohort studies have followed children with prenatal cannabis exposure into adolescence and beyond:

The Ottawa Prenatal Prospective Study (OPPS), led by Peter Fried starting in 1978, has followed participants for decades. Fried’s team found no significant effects on global IQ but reported subtle deficits in executive function, including sustained attention, visual memory, and impulse control. These effects were most pronounced in children whose mothers used cannabis heavily (daily or near-daily) during the first trimester.

The Maternal Health Practices and Child Development (MHPCD) study, led by Nancy Day and Gale Richardson at the University of Pittsburgh, found similar patterns: no gross cognitive deficits but measurable effects on attention, problem-solving, and behavioral regulation that persisted into the teenage years.

A 2020 analysis of data from the Adolescent Brain Cognitive Development (ABCD) Study by Paul et al. in JAMA Psychiatry, which included over 11,000 children aged 9-10, found that prenatal cannabis exposure was associated with increased psychopathology scores, including attention problems and externalizing behaviors. However, the effect sizes were small, and the authors emphasized that the clinical significance of these statistical differences remains unclear.

The pattern across these studies is consistent but subtle: heavy prenatal cannabis exposure appears to be associated with modest disruptions to executive function and behavioral regulation, not catastrophic cognitive damage. Whether light or occasional use produces meaningful effects remains genuinely unknown.

NICU Admission

Several studies have found higher rates of neonatal intensive care unit (NICU) admission among infants born to cannabis-using mothers. A 2019 study by Corsi et al. in JAMA examining over 660,000 Ontario births found that cannabis use was associated with increased odds of NICU admission (aOR 1.40). But this association is difficult to interpret because NICU admission decisions involve clinical judgment and may be influenced by the mother’s disclosed substance use history.

Stillbirth

The data on cannabis and stillbirth is limited and conflicting. A 2013 meta-analysis by Varner et al. in JAMA found no statistically significant association between cannabis use and stillbirth. However, a 2020 cohort study from the Stillbirth Collaborative Research Network found a weak association that was not statistically significant after full adjustment. Currently, there is insufficient evidence to draw conclusions about cannabis and stillbirth risk.

Placental Abruption

Some studies have reported an association between cannabis use and placental abruption, a potentially life-threatening condition where the placenta separates from the uterine wall before delivery. A 2014 study by Saurel-Cubizolles et al. in Placenta found an elevated risk, but the number of exposed cases was small. The evidence remains limited and inconsistent.

The Dose-Response Question

One of the most critical gaps in the literature is the near-total absence of dose-response data. Most studies categorize cannabis use as binary (any use vs. no use) or at best distinguish between occasional and daily use. We have almost no data on what specific doses of THC might be relevant for fetal effects.

This matters enormously because the THC content of cannabis has changed dramatically over the past three decades. The average THC concentration in confiscated cannabis rose from approximately 4% in 1995 to over 15% by 2021, according to data from the University of Mississippi’s Potency Monitoring Program. Concentrates can exceed 80% THC. A woman using low-dose edibles may be getting a fraction of the THC exposure of someone smoking high-potency flower multiple times daily, yet most research treats them identically.

Breastfeeding: A Separate Consideration

THC is secreted into breast milk. A 2018 study by Bertrand et al. in Pediatrics detected THC in 63% of breast milk samples from cannabis-using mothers, with measurable levels persisting for up to six days after last use. THC is lipophilic, and breast milk is high in fat, making it an efficient vehicle for THC transfer.

What this means for infant development is less clear. The amount of THC transferred via breastfeeding is substantially lower than prenatal exposure, and infant metabolism of cannabinoids is poorly characterized. A 2023 systematic review by Koren et al. in Journal of Human Lactation found no consistent evidence of short-term harm from cannabis exposure through breast milk, but acknowledged that long-term studies simply do not exist.

The American Academy of Pediatrics (AAP) and the American College of Obstetricians and Gynecologists (ACOG) both recommend against cannabis use while breastfeeding, citing the precautionary principle. This is a reasonable position given the data gaps, but it should be understood as a precautionary recommendation rather than one backed by strong evidence of demonstrated harm.

Beyond the medical considerations, pregnant women who use cannabis face legal risks that vary dramatically by state. Some states have pursued aggressive interventions:

States with punitive policies: Alabama, Tennessee, and South Carolina have prosecuted women for substance use during pregnancy under child endangerment or chemical endangerment statutes. In Alabama, the chemical endangerment law has been applied to cannabis use, with women facing felony charges carrying up to 10 years in prison.

States with mandatory reporting: In approximately 25 states, healthcare providers are required to report suspected substance use during pregnancy to child protective services. Positive drug screens at birth can trigger CPS investigations, removal of infants, and court-ordered treatment programs regardless of the substance involved.

States with protective approaches: New York, California, and Colorado have taken a more public health-oriented approach, generally treating prenatal substance use as a medical rather than criminal matter. California’s AB 2348 specifically prohibits using a positive drug test alone as evidence of child abuse or neglect.

This legal patchwork means that a woman’s decision to use cannabis during pregnancy carries vastly different legal consequences depending on her zip code. Some reproductive health advocates argue that punitive approaches deter women from seeking prenatal care, ultimately harming the outcomes they are meant to protect.

The Confounding Factor Problem

Nearly every researcher studying cannabis and pregnancy will acknowledge the same frustration: confounding variables make definitive causal conclusions nearly impossible with observational data.

Women who use cannabis during pregnancy are statistically more likely to also use tobacco, have lower incomes, experience higher rates of stress and anxiety, have less access to prenatal care, and report higher rates of domestic violence. Each of these factors independently affects pregnancy outcomes. Statistical adjustment can reduce confounding but cannot eliminate it entirely, particularly for unmeasured or poorly measured confounders.

The gold standard for establishing causation, the randomized controlled trial, is obviously unethical in this context. No IRB would approve randomly assigning pregnant women to use THC. This means we are fundamentally limited to observational evidence, natural experiments, and animal models.

Animal studies, primarily in rats and mice, have shown more pronounced effects of prenatal cannabinoid exposure on offspring neurodevelopment. A 2018 study by de Salas-Quiroga et al. in Nature Medicine found that prenatal THC exposure in mice altered cortical development through epigenetic mechanisms. But rodent models have important limitations: their gestational neurodevelopmental timelines differ from humans, they metabolize cannabinoids differently, and the doses used often far exceed typical human consumption.

What Medical Organizations Say

The major medical organizations have converged on a precautionary message:

  • ACOG recommends discontinuing cannabis use during pregnancy and lactation, and advises physicians to screen for cannabis use and counsel patients about potential risks.
  • AAP recommends against cannabis use during pregnancy and breastfeeding.
  • WHO identifies prenatal cannabis exposure as a potential risk factor for low birth weight and cognitive effects.
  • Society of Obstetricians and Gynaecologists of Canada (SOGC) recommends against cannabis use in pregnancy, noting the evidence for potential harm to fetal neurodevelopment.

Notably, all of these recommendations use language like “potential risk” and “recommend against” rather than stating that harm has been definitively established. This language reflects the genuine uncertainty in the evidence base.

What Honest Counsel Looks Like

Given the current state of the evidence, the most responsible summary looks something like this:

What appears reasonably established: THC crosses the placenta and reaches the fetal brain. Heavy, daily cannabis use during pregnancy is associated with lower birth weight and subtle effects on executive function in offspring. These associations persist across multiple large studies even after adjustment for major confounders.

What remains uncertain: Whether light or occasional use carries meaningful risk. Whether CBD poses its own risks. What dose thresholds, if any, separate negligible from significant effects. Whether modern high-potency cannabis changes the risk calculus compared to older studies.

What is not supported by the evidence: Claims that any amount of prenatal cannabis exposure causes severe developmental damage. Claims that cannabis use during pregnancy is proven safe. Comparisons to fetal alcohol syndrome, which involves a fundamentally different mechanism and much stronger evidence base.

The evidence does not justify panic, but it also does not support complacency. For women who can stop using cannabis during pregnancy without significant hardship, stopping is the most conservative and defensible choice. For women who are using cannabis to manage severe hyperemesis gravidarum or other conditions where pharmaceutical alternatives have failed or carry their own fetal risks, the decision becomes a genuine risk-benefit calculation that should involve a trusted healthcare provider.

What the evidence definitively does not support is criminalizing pregnant women, shaming them, or using unreliable drug screens to separate families. The history of the “crack baby” moral panic, which was driven more by racism and politics than by science and whose predicted catastrophic outcomes never materialized, should serve as a cautionary tale about the dangers of conflating social anxiety with medical evidence.

The Path Forward for Research

The single most important thing that could happen for this field is better research. The barriers are significant but not insurmountable:

Prospective cohort studies that carefully characterize dose, frequency, route of administration, THC/CBD ratios, and timing of exposure relative to gestational milestones would dramatically improve our understanding. The ongoing HEALthy Brain and Child Development (HBCD) Study, funded by the NIH, is designed to do exactly this.

Biomarker development for quantifying fetal cannabinoid exposure would help move beyond self-report, which is known to underestimate actual use. Meconium and umbilical cord tissue analysis show promise but need standardization.

Epigenetic studies examining whether prenatal cannabis exposure creates heritable changes in gene expression could reveal effects that are not apparent from behavioral testing alone. Early work by Szutorisz et al. in Neuropsychopharmacology has found cross-generational epigenetic effects in animal models, but human data is scarce.

Until that research arrives, we owe pregnant women something better than scare tactics and something more honest than reassurance. We owe them the evidence as it actually stands: incomplete, suggestive of caution, and desperately in need of better data.