Cannabis use during pregnancy is linked to a moderately higher risk of miscarriage, though researchers still debate how much of that risk comes from THC itself versus the other factors that tend to travel alongside it. A 2022 meta-analysis pooling data from multiple studies found that cannabis users had roughly 28 percent higher odds of pregnancy loss at any gestational age, and 39 percent higher odds of loss before 22 weeks, compared with nonusers. The picture is more complicated than a single number, though, because the body’s own cannabis-like signaling system plays an unexpectedly central role in whether an embryo successfully implants in the first place.
What the Population-Level Evidence Shows
The most comprehensive look at this question comes from a systematic review and meta-analysis published in Fertility and Sterility. Pooling the available studies, researchers found that cannabis use was associated with 28 percent higher odds of pregnancy loss overall. When the analysis was restricted to earlier losses, those occurring before 22 weeks of gestation, the association was stronger: cannabis users had 39 percent higher odds of losing a pregnancy compared with nonusers.1Fertility and Sterility. Cannabis use and pregnancy loss: a systematic review and meta-analysis That is a meaningful signal, but it is worth putting in context. A 28 to 39 percent increase in relative odds sounds alarming, but the baseline risk of miscarriage for any given pregnancy is already around 10 to 20 percent. So the absolute increase in risk, while real, is smaller than the relative numbers suggest on first glance.
One limitation that runs through nearly all studies on this topic is that they are observational. Nobody is going to randomize pregnant people to smoke cannabis or not, so researchers have to rely on comparing women who report using cannabis with those who do not. That leaves room for confounding: cannabis users may differ from nonusers in other ways, such as stress levels, income, tobacco use, or access to prenatal care, and those differences could partly explain the results.
How THC Interferes with Implantation
Your body produces its own cannabis-like molecules called endocannabinoids, and they are critical for the earliest stages of pregnancy. The endocannabinoid system helps govern embryo development, transport through the fallopian tube, implantation into the uterine wall, and the initial formation of the placenta.2PubMed Central. Endocannabinoid System in Pregnancy Maintenance and Labor: A Mini-Review Two of the body’s main endocannabinoids, anandamide and 2-AG, are essential for the uterine lining to prepare for a pregnancy and for the embryo to implant properly.3PubMed. Role for endocannabinoids in early pregnancy: recent advances and the effects of cannabis use
The key insight from laboratory research is that anandamide has to stay within a very narrow concentration range for implantation to succeed. At low concentrations, it activates signaling pathways in the embryo’s outer cell layer that make the embryo ready to attach to the uterine wall. At higher concentrations, just a fourfold increase in animal models, it instead suppresses different signaling channels and blocks the embryo from implanting.4PubMed Central. Differential G protein-coupled cannabinoid receptor signaling by anandamide directs blastocyst activation for implantation That same research noted that elevated anandamide levels have been linked to spontaneous pregnancy loss in women. When THC from cannabis enters the picture, it acts on the same receptors that anandamide uses, effectively flooding a system designed to operate within tight margins.
Think of it like a thermostat that needs to sit at exactly the right temperature. The body carefully dials its endocannabinoid levels up and down at different stages of early pregnancy. Introducing an outside source of cannabinoids disrupts that calibration, potentially tipping the balance toward failed implantation or early loss.
Placental Damage and Blood Vessel Problems
Beyond implantation, THC appears to harm the placenta itself, particularly its network of blood vessels. A study examining placentas from women who smoked cannabis at least once a month during pregnancy (without concurrent tobacco or alcohol use) found that their placental blood vessel networks were narrower and less developed than those of nonusers.5Toxicology Letters. RhoA/MLC signaling pathway is involved in Δ⁹-tetrahydrocannabinol-impaired placental angiogenesis A healthy placenta needs a dense, well-branched system of blood vessels to shuttle oxygen and nutrients from mother to fetus. When that network is stunted, the placenta cannot do its job properly.
Animal research supports and extends these findings. In a rat study, THC exposure during pregnancy led to specific defects in the labyrinth zone of the placenta, the region where maternal and fetal blood come closest together to exchange gases and nutrients. Fetal blood space was reduced while maternal blood space was expanded, disrupting the delicate architecture the placenta needs for efficient exchange. The pregnancies also resulted in symmetrical fetal growth restriction, meaning the fetuses were uniformly smaller rather than just having one undersized organ.6Scientific Reports. Δ9-tetrahydrocannabinol exposure during rat pregnancy leads to symmetrical fetal growth restriction and labyrinth-specific vascular defects in the placenta
Trophoblast cells, which are the cells that burrow into the uterine wall and build the maternal-fetal interface, also seem to be affected. In animal models where the CB1 cannabinoid receptor was knocked out entirely, trophoblast stem cells were less able to invade properly, suggesting that the endocannabinoid system’s involvement in placental construction is not optional, it is structural.7PubMed Central. Conventional Cigarettes, E-Cigarettes, and Marijuana Use In Pregnancy Understanding the Placental Biology of Tobacco Smoke, Nicotine, and Marijuana (THC) Exposures During Pregnancy
What Miscarriage Tissue Reveals About Endocannabinoid Disruption
Some of the most striking evidence comes from examining placental tissue after spontaneous miscarriages. Researchers compared first-trimester placentas from women who miscarried with placentas from women who had elective terminations at a similar gestational age. The miscarriage placentas showed a distinctive pattern: high levels of the CB1 cannabinoid receptor and low or absent levels of FAAH, the enzyme responsible for breaking down anandamide.8Placenta. Endocannabinoid System in First Trimester Placenta: Low FAAH and High CB1 Expression Characterize Spontaneous Miscarriage
This matters because FAAH acts as a cleanup crew for anandamide. When FAAH is scarce, anandamide accumulates. When CB1 receptors are simultaneously overexpressed, the tissue becomes hyper-responsive to whatever endocannabinoids are present. The result is a placenta that is, in effect, drowning in its own cannabinoid signaling. This finding does not prove that cannabis use caused those miscarriages, since the altered enzyme and receptor levels could be a consequence of other problems rather than a cause. But it does show that the endocannabinoid system is not a bystander in pregnancy loss: when that system is out of balance, miscarriage follows.
Does It Matter How Much or How Often You Use?
The evidence points toward a dose-response relationship. A study published in JAMA that measured THC metabolite levels in urine, rather than relying on what women reported, found that higher cumulative THC exposure across pregnancy was associated with a greater probability of adverse placental outcomes. Women whose cumulative metabolite levels crossed a high threshold had roughly 60 percent higher odds of the composite adverse outcome in adjusted models.9JAMA. Cannabis Exposure and Adverse Pregnancy Outcomes Related to Placental Function That finding is important because it shifts the conversation from “did she use or not” toward “how much THC actually reached her system,” which is a more biologically meaningful question.
Modern cannabis products are substantially more potent than what was available even a decade ago, which means that the studies from the 1980s and 1990s that found weak or no associations may simply have been studying lower exposures. A few puffs of today’s high-THC flower or a hit from a concentrate pen delivers far more THC than the same behavior would have in prior decades, and the research has not fully caught up to these changes in the market.
Does the Father’s Cannabis Use Play a Role?
This is a question that tends to get overlooked, but the early evidence is worth paying attention to. A prospective cohort study found that when male partners used marijuana at least once a week before conception, the couple’s risk of spontaneous miscarriage roughly doubled compared with couples where the male partner did not use, after adjusting for other factors.10PubMed Central. Male Preconception Marijuana Use and Spontaneous Abortion: A Prospective Cohort Study Less-than-weekly use did not show a clear association. A broader review of male substance use and reproductive outcomes echoed this, noting that preconception paternal cannabis use has been linked to increased rates of pregnancy loss alongside other outcomes like lower birth weight.11Nature Reviews Urology. Influence of substance use on male reproductive health and offspring outcomes
The mechanism here is thought to involve sperm. THC can alter sperm DNA methylation patterns, sperm motility, and sperm count. If the genetic or epigenetic material packaged in a sperm cell is compromised, the resulting embryo may be less viable from the start, even if the pregnant partner never used cannabis herself. This is still a relatively new area of investigation, and the evidence is not as deep as it is for maternal use, but it is a genuine finding from well-designed prospective research rather than speculation.
The Underreporting Problem
A fundamental challenge in studying cannabis and pregnancy is that many people do not disclose their use. A scoping review of studies comparing self-reported cannabis use during pregnancy with biological testing found that self-report was largely unreliable: biochemical tests consistently detected higher rates of prenatal cannabis use than women admitted to.12PubMed Central. Validity of self-report measures of cannabis use compared to biological samples among women of reproductive age: a scoping review In one study in Northern California, urine toxicology identified nearly twice as many cannabis users as self-report did, and the sensitivity of self-report was only about 34 percent, meaning two out of three users who tested positive had not disclosed their use.13PubMed Central. Validity of Self-Reported Cannabis Use Among Pregnant Females in Northern California
In a large population-based cohort from the Netherlands, the picture was similarly murky. Only about a third of women who reported using cannabis during pregnancy also tested positive on urine screening, and positive urine results turned up among women who reported using only before pregnancy or who gave no information about use at all.14European Addiction Research. Agreement between Maternal Cannabis Use during Pregnancy according to Self-Report and Urinalysis in a Population-Based Cohort: The Generation R Study
This underreporting matters in a specific way for the research. If many cannabis users in studies are miscategorized as nonusers, the true association between cannabis and miscarriage is probably being diluted. The real risk could be higher than what the published numbers show, because the “unexposed” group in most studies quietly contains some exposed women. Researchers who have grappled with this problem generally agree that a combination of self-report and biochemical testing is needed to get accurate data, but most large studies still rely on self-report alone.
Separating Cannabis from Tobacco and Other Confounders
One of the most common pushbacks on this research is that people who smoke cannabis during pregnancy may also smoke tobacco, drink alcohol, experience more stress, or have other characteristics that independently raise miscarriage risk. This is a legitimate concern, but some studies have tried to address it directly. In a study of infertility patients undergoing assisted reproduction, researchers noted that the overlap between cannabis and tobacco use in their population was actually low: only about a quarter of female cannabis users also smoked tobacco. Additionally, the pattern of associations with treatment outcomes was different for tobacco and cannabis, suggesting the two substances were not simply proxying for each other.15Human Reproduction. Marijuana smoking and outcomes of infertility treatment with assisted reproductive technologies
Stress and depression are another entangled factor. Some women use cannabis specifically to cope with prenatal stress or mood problems. A study tracking cannabis use, stress, and depression trajectories across trimesters found that mental health history and current stress levels were associated with cannabis use patterns, meaning the two are deeply intertwined.16medRxiv. Coping with Cannabis During Pregnancy: Trajectories of Depression, Stress, and Cannabis Use across the Prenatal Period Stress itself is a known risk factor for miscarriage, so disentangling whether THC, stress, or the combination is driving the risk remains genuinely difficult. This does not mean cannabis is off the hook; it means the honest answer is that the biological evidence for a direct effect is compelling, but the population-level studies cannot cleanly isolate it from other risk factors.
Contaminants Add a Separate Layer of Risk
Cannabis is not just THC. Depending on how it was grown and processed, the plant material or concentrate may carry microbes, heavy metals, and pesticide residues. A review of cannabis contaminants noted that their direct human toxicity is poorly quantified but that documented effects include infection risk, potential carcinogenicity, and reproductive and developmental impacts.17PubMed Central. Cannabis contaminants: sources, distribution, human toxicity and pharmacologic effects In states and countries with regulated markets, products go through some level of testing, but unregulated or illicit-market cannabis has no such oversight. For someone who is pregnant, even a small burden of lead or cadmium from contaminated flower is an exposure that would not exist without the cannabis use.
Smoking as a delivery method adds combustion byproducts to the mix. While some people assume that edibles or vaporizers avoid these risks, the THC still reaches the body and the endocannabinoid system. The combustion-related toxins are an additional concern on top of the THC itself, not a replacement for the underlying question of cannabinoid exposure.
What Medical Organizations Recommend
The American College of Obstetricians and Gynecologists, the largest professional body for OB-GYNs in the United States, updated its clinical consensus on cannabis and pregnancy in 2025. The guidance calls for obstetric professionals to screen all patients for cannabis use and to use evidence-based strategies to help patients reduce or stop use during pregnancy and lactation.18PubMed. ACOG Clinical Consensus No. 10: Cannabis Use During Pregnancy and Lactation The framing is cautious rather than absolutist: the evidence is strong enough that the recommendation is to avoid cannabis, but clinicians are encouraged to have nonjudgmental conversations rather than deliver ultimatums, partly because shaming patients tends to reduce disclosure rather than reduce use.
That stance reflects the overall evidence landscape. There is a plausible biological mechanism for harm, the endocannabinoid system’s role in implantation and placentation is not speculative. There is a consistent observational signal across multiple studies. There is a dose-response pattern. And there is no identified safe threshold. What is missing is a definitive randomized trial, which will never exist for ethical reasons. In the absence of that trial, the medical consensus leans toward caution, recommending that people who are pregnant or planning to become pregnant avoid cannabis use entirely.
CBD Products and the “It’s Not THC” Assumption
A common follow-up question is whether CBD products, which contain little or no THC, are a safe alternative during pregnancy. The short answer is that nobody knows, because almost no research has specifically tested CBD’s effects on human pregnancy outcomes. CBD does interact with the endocannabinoid system, albeit differently than THC. It influences cannabinoid receptor activity indirectly and affects the breakdown of anandamide, the same molecule whose precise concentration is critical for implantation. Phytocannabinoids broadly, not just THC, may have negative effects on fertility, pregnancy, and fetal health through their action on cannabinoid receptors.3PubMed. Role for endocannabinoids in early pregnancy: recent advances and the effects of cannabis use
CBD products also come with their own regulatory ambiguity. Many products on the market contain more THC than their labels claim, and they may carry the same contaminant risks as other cannabis products. Until specific safety data exist for CBD in pregnancy, the assumption that it is harmless is exactly that: an assumption, not a conclusion supported by evidence. ACOG’s guidance does not carve out an exception for CBD, treating it as part of the broader category of cannabis products to avoid during pregnancy.