OBJECTIVE: To evaluate antidepressant switching rates and associated dosing characteristics in patients undergoing therapy with one of the selective serotonin reuptake inhibitors (SSRls)-fluoxetine, paroxetine, or sertraline. DESIGN: A longitudinal, retrospective drug utilization analysis of administrative prescription claims for patients who initiated therapy between 2/1/95 and 4/30/95. Switching must have occurred during the initial episode of antidepressant care. Study measures included rates of switching to any antidepressant, rates of switching to another SSRI, and daily doses for the first and last prescriptions of the first and second SSRls. PATIENTS: 16,582 study patients, 11HJ4 years of age, which had used no antidepressant or other psychotropic medication in the six months prior to initiating SSRI therapy, and had no other psychotropic medication use in the 12 months following therapy. RESULTS: Rates of switching to any other antidepressant were significantly different for fluoxetine (4.5%), paroxetine (10.0%), and sertraline (6.0%). Rates of switching to another SSRI were 51.2% for fluoxetine, 58.5%' for paroxetine, and 49.0% for sertraline. In these patients, switching tended to occur from the recommended minimum dose of the first SSRI (71.2%) and to the recommended minimum dose of the second SSRI (76.5%). CONCLUSION: The differences in switching rates among fluoxetine, paroxetine, and sertraline may reflect product differences in clinical practice. It appears from these findings that, overall, physicians viewed recommended dose minimums for SSRls as therapeutically interchangeable, although in a minority of patients, it appears sertraline use may necessitate higher dosing relative to fluoxetine or paroxetine.