BACKGROUND: The goal of this study was to examine if in utero cannabis exposure predicted reduced birth size and if these effects were evident in specific growth parameters as early as the second trimester. METHODS: Eligible women had an initial prenatal visit between January 1, 2010, and March 31, 2020, completed an anatomy ultrasound between 18–24 weeks’ gestation, and had no self-reported alcohol, tobacco, or other biochemically verified drug use. The two primary study groups were cannabis users (n = 109) identified through self-report and urine toxicology screens, and a randomly selected control group of non-substance users (n = 171). Medical records were manually reviewed for background and medical information, anatomy ultrasound results, and birth size parameters. RESULTS: After controlling for significant confounders, regression results indicated significant (p < .05) deficits in birth weight associated with cannabis exposure, with a trend for increasing weight effects beginning in the second trimester. A significant decrease in head circumference was evident as early as the second trimester, with differences remaining significant until birth. Significant overall length and specific bone length deficits were not predicted by cannabis exposure, at birth or earlier in gestation, after control for confounding. CONCLUSIONS: Cannabis exposure predicted growth deficits at birth, with some effects evident as early as the second trimester. The biggest and earliest effects were seen for cranial size, with an adjusted deficit of more than 14 percentile points by birth. Overall weight was not impacted until at or near delivery.