Objective: Previous work in adults and youths has suggested that juvenile onset bipolar disorder (BPD) is associated with an elevated risk of substance use disorders (SUD). Considering the public health importance of this issue, the authors now report on a controlled study of adolescents with and without BPD to evaluate the risk of SUD. Method: Probands with DSM-IV BPD (n = 57, mean age [+ or -] SD = 13.3 [+ or -] 2.4 years) and without DSM-IV BPD (n = 46, 13.6 [+ or -] 2.2 years) were studied. Structured psychiatric interviews and multiple measures of SUD were collected. Results: Bipolar disorder was associated with a highly significant risk factor for SUD (32% versus 7%, Z = 2.9, p = .004) that was not accounted for by conduct disorder (adjusted odds ratio = 5.4, p = .018). Adolescent-onset BPD ([greater than or equal to]13 years) was associated with a higher risk of SUD compared with those with child-onset BPD ([[chi square].sub.1] = 9.3, p = .002). Conclusions: These findings strongly indicate that BPD, especially adolescent onset, is a significant risk factor for SUD independently of conduct disorder.