Recent media reports have raised concerns about the use of second-generation antipsychotics to treat children. Child psychiatrists and others, worried in turn that these reports will discourage parents from seeking help, have urged the public not to take research findings out of context and to consider the potential dangers of not treating serious symptoms. These paired concerns are reflected in this month's issue. The lead article reports that 41% of children in a state Medicaid population who received an antipsychotic had diagnoses for which there was no evidence to support its use. Analyses of claims data by Prathamesh Pathak, M.S., B.Pharm., and colleagues determined that 11,700 children received a new prescription for a second-generation antipsychotic between 2001 and 2005. As the authors note, none of these agents were approved for pediatric use until 2006, although findings from more than 80 clinical studies of pediatric use of these agents for a range of diagnoses were available to prescribers before the end of 2005. The authors classified these findings as providing strong, plausible, weak, or no evidence of effectiveness. Their review of diagnoses given to the 11,700 children found that 4,834 did not have a disorder for which published evidence supported the agent's use (page 123). In Taking Issue, Bonnie T. Zima, M.D., M.P.H., points out that assigning weight to scientific evidence requires careful deliberation, and she calls for scrutiny of the study's findings because they are liable to be misinterpreted (page 107). In a study of second-generation antipsychotic use by nearly 309,000 elderly nursing home residents in 2004, Pravin Kamble, B.Pharm., M.S., and colleagues found that 86% received them for indications that were off label during the study period and 57% received them for an evidence-based use (page 130).