Maternal cocaine use during pregnancy can affect the infant directly through toxic effects or indirectly through cocaine's influence on maternal psychological status. We followed 160 cocaine exposed and 56 nonexposed infants and their mothers identified at birth through interview and/or urine screen. Although cocaine exposure defined the groups, infant exposure to alcohol, marijuana, and tobacco was allowed to vary. Infants were 99% African American and poor. All mothers completed the Brief Symptom Inventory (BSI) and infants were given the Bayley Scales of Mental (MDI) and Motor (PDI) Development at a mean corrected age of 17 +/- 8 months. Both MDIs (94 +/- 17 vs. 103 +/- 16) and PDIs (101 +/- 16 vs. 108 +/- 12) were lower for cocaine exposed infants. Psychological distress was greater in cocaine using mothers. Hierarchical multiple regression was used to assess the relative effects of gestational age, maternal psychological distress, and cocaine and polydrug exposure on infant outcomes. Both psychological distress and cocaine and alcohol exposure predicted lower MDIs after controlling for prematurity. Neither psychological distress nor alcohol exposure predicted motor outcome, while cocaine had a significant effect. Tobacco and marijuana exposure were unrelated to outcome. These findings provide further support for direct effects of cocaine and alcohol on infant development as well as highlight the need for studies to document maternal psychological factors, which may increase child risk for poorer outcomes.