Objectives: Public Law (P.L.) 110-351, the "Fostering Connections to Success Act," calls for state child welfare agencies to partner with Medicaid and pediatric experts to provide planning and oversight regarding the provision of health and mental health services, including medication, to children in state custody. Recent reports, media cases, and class action lawsuits suggest over-use of psychotropic medications to address the behavioral needs of children in the child welfare system. We examined geographic variability in psychotropic medication use across US child welfare agency catchment areas to determine how rates of psychotropic medication use vary in relation to child, community, child welfare, and health system-level factors. Methods: Cross-sectional analysis of Wave 1 data for the 92 child welfare catchment areas participating in the National Survey of Child and Adolescent Well-being, a random probability sample of 2,504 children ages 2-15 years undergoing investigation for abuse and neglect. We employed multilevel regression modeling to examine the impact of catchment-level variables on medication use, controlling for child-level variables. Results: Fifteen percent of children reported taking psychotropic medications. Rates of medication use across catchment areas ranged widely from 0 to 40%, a 40-fold difference. On multi-level logistic regression modeling, older age (p less than 0.001), male gender (p less than 0.001), emotional and behavioral problems (p less than 0.001), and insurance (p = 0.05) were associated with psychotropic medication use at the child-level. At the catchment-level, stressful environment within the child welfare system was negatively associated with medication use. No other catchment-level variables examined were found to explain use. Conclusions: Striking disparities in medication use exist across catchment areas in this national sample. Of the catchment variables examined, only stressful environment was related to medication use. Practice implications: These findings highlight significant geographic variation in medication use that most likely reflect both under-use and over-reliance on psychotropic medication. The link between child welfare environment and medication use suggests the influence of systemic, as opposed to clinical, causes of variation in medication use. This requires greater implementation of organizational processes governing quality of care for this highly vulnerable population. (Contains 3 tables and 1 figure.)