OBJECTIVE: To investigate a hypothesized link between antidepressant (AD) drug use in pediatric patients with major depressive disorder (MDD) and an increased rate of suicide attempts. METHODS: A retrospective cohort study was conducted using claims data from the PharMetrics Integrated Outcomes Database from 1997-2002. Adolescents (aged 12-18 at time of MDD diagnosis) who were continuously enrolled in a managed health plan; had no prior diagnosis of MDD, AD use, or psychotherapy for one year; and had at least 6 months of follow-up data were included. Cox Proportional Hazard models were created to evaluate the multivariate relationship between AD use and time to suicide attempt, controlling for calendar year, demographic and clinical characteristics, and patient compliance. AD use was measured using pharmacy claims; suicide attempts were measured using ICD-9 codes in any health care setting. RESULTS: A total of 15,826 subjects (63% of whom were female) met inclusion criteria. AD treatment with any class of drugs did not increase the risk of suicide attempt (all Hazard Ratios NS). No statistically significant differences were found between more- and less-activating agents. AD treatment for less than 6 months was associated with increased risk of suicide attempt: subjects taking AD for 1-55 days (HR = 2.93, 95% CI = 1.34-6.39) or 56 - 179 days (HR = 3.26, 95% CI = 1.81-5.87) were more likely than subjects taking AD for 180 days or more to attempt suicide. Adolescent females, those receiving psychotherapy within 90 days of MDD diagnosis, those with diagnosed substance abuse, schizophrenia, or another mental health disorder, those with more chronic diseases, and those in the Midwest and West were independently at greater risk for suicide attempts. CONCLUSIONS: Antidepressant drug use had no independent effect on the likelihood of suicide attempt for a large cohort of adolescents across the U.S. The relationship between AD drug use and suicidality is complex and requires further investigation.