OBJECTIVES: The objectives of the study were to examine persistence and factors associated with persistence to antidepressant medications in depressed children and adolescents and to evaluate health care expenditures across persistent patients and non-persistent patients. METHODS: Medical Expenditure Panel Survey (MEPS) data from the years 2002-2007 were used for the analysis. The study sample consisted of depressed children and adolescents taking an antidepressant. Multiple logistic regression analysis was conducted to examine factors associated with persistence, after controlling for various independent variables like the predisposing, enabling and need factors chosen on the basis of Anderson Behavioral Model, and controls for year and timeframe. Linear regression models, containing log of prescription drug expenditures and ambulatory expenditures as the dependent variables, were used to evaluate the expenditures. Unweighted statistical analysis was performed due to small sample cell size. RESULTS: Analytical sample consisted of 281 children and adolescents using antidepressants. Approximately 53% of the children and adolescents initiated on the antidepressant therapy were not persistent. Social limitation was found to be a significant predictor increasing persistence. Switching was prevalent in 16% of the children and adolescents. Analysis of expenditures revealed that being persistent led to a 2.5 times increase in the prescription drug expenditures, with no impact on ambulatory care expenditures. CONCLUSIONS: Over half of the patients on antidepressant therapy were not persistent. Although persistence of antidepressant therapy did not translate into reduction in ambulatory expenditures, it may play a role in clinical and humanistic outcomes.