Context: The relation between use of antidepressants, especially selective serotonin reuptake inhibitors (SSRIs), and suicidal ideation and behaviors has received considerable public attention recently. The use of such drugs among teenagers has been of particular concern. Objective: To estimate the relative risks (RRs) of nonfatal suicidal behavior in patients starting treatment with 1 of 3 antidepressant drugs compared with patients starting treatment with dothiepin. Design and setting: Matched case-control study of patients treated in UK general practices using the UK General Practice Research Database for 1993—1999. Participants: The base population included 159,810 users of the 4 antidepressant drugs. Participants could have used only 1 of these antidepressants and had to have received at least 1 prescription for the study antidepressant within 90 days before their index date (the date of suicidal behavior or ideation for cases and the same date for matched controls). Main outcome measures: Frequency of first-time exposure to amitriptyline, fluoxetine, paroxetine, and dothiepin of patients with a recorded diagnosis of first-time nonfatal suicidal behavior or suicide compared with comparable patients who did not exhibit suicidal behavior. Results: After controlling for age, sex, calendar time, and time from first antidepressant prescription to the onset of suicidal behavior, the relative risks for newly diagnosed nonfatal suicidal behavior in 555 cases and 2062 controls were 0.83 (95% confidence interval, [CI] 0.61-1.13) for amitriptyline, 1.16 (95% CI, 0.90-1.50) for fluoxetine, and 1.29 (95% CI, 0.97-1.70) for paroxetine compared with those using dothiepin. The RR for suicidal behavior among patients first prescribed an antidepressant within 1 to 9 days before their index date was 4.07 (95% CI, 2.89-5.74) compared with patients who were first prescribed an antidepressant 90 days or more before their index date. Time since first antidepressant prescription was not, however, a confounder of the relation between specific antidepressants and suicidal behavior since its relation to suicidal behavior was not materially different among users of the 4 study drugs. Similarly for fatal suicide, the RR among patients who were first prescribed an antidepressant within 1 to 9 days before their index date was 38.0 (95% CI, 6.2-231) compared with those who were first prescribed an antidepressant 90 days or more before their index date. There were no significant associations between the use of a particular study antidepressant and the risk of suicide. Conclusions: The risk of suicidal behavior after starting antidepressant treatment is similar among users of amitriptyline, fluoxetine, and paroxetine compared with the risk among users of dothiepin. The risk of suicidal behavior is increased in the first month after starting antidepressants, especially during the first 1 to 9 days. A possible small increase in risk (bordering statistical significance) among those starting the newest antidepressant, paroxetine, is of a magnitude that could readily be due to uncontrolled confounding by severity of depression. Based on limited information, we also conclude that there is no substantial difference in effect of the 4 drugs on people aged 10 to 19 years.