This study examined the role of maternal psychopathology and maternal warmth as mediators of the association between prenatal cocaine and other substance exposure and toddler behavior problems. It was also hypothesized that infant cortisol reactivity and environmental risk may moderate these associations. Participants were 220 caregiver-infant dyads (119 cocaine exposed, 101 not cocaine exposed; 49% boys). Mother-infant dyads were recruited at delivery with assessments at 4-8 weeks and 7, 13, and 18 months of child ages. Results yielded no direct associations between prenatal cocaine/other substance exposure and toddler behavior problems, but significant indirect associations between prenatal cigarette/alcohol exposure and toddler behavior problems at 18 months. With regard to moderation, results indicated an indirect association between prenatal cocaine exposure and toddler behavior problems via lower maternal warmth for children with higher, but not lower, cortisol reactivity at 7 months. Results suggest potential pathways to toddler behavior problems among children at high biological risk.