Studies have investigated the potential effects of prenatal cocaine exposure (CE) on children's development. However, few studies have examined predictors of resilient outcomes in this population. We examined vagal tone (VT) as a resilience factor in prenatal CE. Utilizing data from the Maternal Lifestyle Study, a cumulative risk index was derived for children with and without prenatal CE. Presence of CE and other prenatal drugs was summed with postnatal risks in infancy to yield a 15-item risk index. Preschool cognitive outcomes, problem behaviors, and adaptive behaviors were measured. VT was assessed during an infant exam at 1 month and toy exploration at 36 months. We included children with complete physiologic data (217 CE, 333 non-CE). Children were classified as having consistently high, consistently low, or fluctuating VT at 1 and 36 months. Children were also classified as high versus low risk. High-risk children had lower IQ scores, more problem behaviors, and lower ratings of adaptive behaviors than low-risk children. A significant risk by VT–stability interaction indicated that for high-risk children, those with stable low VT had higher ratings of adaptive behaviors at 36 months. This is consistent with theory linking reduced VT during tasks to adaptive regulation and indicates that such regulatory functioning may serve as a protective factor in prenatal CE.