OBJECTIVE Examined effects of prenatal cocaine exposure (PCE) on tobacco, alcohol, marijuana and cocaine use by age 15. METHODS Adolescent (n=358; 183 PCE, 175 non-prenatally cocaine exposed; NCE) drug use was assessed using urine, hair, and/or blood spot samples and self-report (Youth Risk Behavior Surveillance System; YRBSS) at ages 12 and 15. Logistic regression assessed effects of PCE on drug use controlling for other drug exposures, environment and blood lead levels (BLL). RESULTS Adjusted percentages of drug use (PCE vs. NCE) were: tobacco 35% vs. 26% (p<.04), marijuana 33% vs. 23% (p<.04), alcohol 40% vs. 35% (p<.01), and any drugs 59% vs. 50% (p<.005). PCE adolescents were twice as likely to use tobacco (OR=2.02, 95% CI=1.05-3.90, p<.04), 2.2 times more likely to use alcohol (OR=2.16, 95% CI=1.21-3.87, p<.01) and 1.8 times more likely to use marijuana (OR=1.81, 95% CI=1.02-3.22, p<.04) than NCE adolescents. A race-by-cocaine-exposure interaction (p<.01) indicated PCE non-African American adolescents had greater probability of tobacco use (65%) than NCE non-African American youth (21%). PCE was associated with any drug use (OR=2.16, CI=1.26-3.69, p<.005), while higher BLL predicted alcohol use (p<.001). Violence exposure was a predictor of tobacco (p<.002), marijuana (p<.0007) and any drug (p<.04). CONCLUSIONS PCE and exposure to violence increased the likelihood of tobacco, marijuana or any drug use by age 15, while PCE and higher early BLL predicted alcohol use. Prevention efforts should target high risk groups prior to substance use initiation.