OBJECTIVE The Child and Adolescent First-Episode Psychosis Study (CAFEPS) is a naturalistic longitudinal study of early-onset first psychotic episodes. This report describes the antipsychotic treatment during the first year and compares the most frequently used agents after 6 months. METHODS Participants were 110 patients, aged 9-17 years, with a first psychotic episode attended consecutively at six different centers. The Positive and Negative Symptom Scale (PANSS), Clinical Global Impressions (CGI), Disability Assessment Schedule (DAS), and Global Assessment of Function (GAF) scales were administered at baseline and at 6 months and the Udvalg for Kliniske Undersøgelser (UKU) Side Effects Rating Scale only at 6 months. RESULTS Diagnoses at baseline were 38.2% psychotic disorder not otherwise specified, 39.1% schizophrenia-type disorder, 11.8% depressive disorder with psychotic symptoms, and 10.9% bipolar disorder, manic episode with psychotic symptoms. The most frequently used antipsychotic agents were risperidone (n = 50), quetiapine (n = 18), and olanzapine (n = 16). Patients who were prescribed olanzapine or quetiapine had more negative and general symptoms. Using the baseline score as covariate, no significant differences were found in the reductions on any scale in patients treated with risperidone, quetiapine, or olanzapine for 6 months. Weight increase was greater with olanzapine than with risperidone (p = 0.020) or quetiapine (p = 0.040). More neurological side effects appeared with risperidone than with olanzapine (p = 0.022). All side effects were mild or moderate. CONCLUSIONS Second-generation antipsychotics, especially risperidone, quetiapine, and olanzapine, are the most used in our context in first psychotic episodes in children and adolescents. These three obtain similar clinical improvement, but differ in their side effects.