OBJECTIVES To examine whether educational mailings to prescribers of psychotropic medications to Medicaid beneficiaries improve patient quality of care and reduce drug costs. METHODS Prescribing behavior was compared to pre-intervention behavior and, where available, to the behavior of prescribers who were not sent mailings. Medicaid pharmacy and eligibility data from April 2003 through September 2007 were used, as well as data on intervention mailings and deviations from clinical guidelines (from which the treatment group was identified). Primary outcome measures included clinical deviations, psychotropic claims, and psychotropic costs as a percentage of total psychotropic claims or costs. Intervention effects were estimated with generalized linear regression models controlling for prescriber-level fixed effects and a time trend, including sensitivity analyses for various prescriber subgroups. In Phase I (March 2004 to November 2005) mailings were sent monthly to qualifying prescribers (297; 221 comparison) and in Phase II (March 2006 to January 2007) mailings were alternated between prescribers to children (89; 82 comparison) and prescribers to adults (145; 354 comparison). RESULTS Impacts were small and generally not statistically significant. Prescribing behavior in Phase I and for prescribers to adults in Phase II was not different from existing trends (p > 0.10 for all outcomes). For prescribers to children, intervention period trends were statistically different from pre-existing ones for all outcomes; however, they were not different when compared with comparison group prescribers. For example, targeted claims as a percentage of total claims for treatment group prescribers were slightly lower than for the comparison group, but the average difference was less than 1 percentage point (p = 0.12). Effects for all outcomes were not significant until 12 months into the intervention. CONCLUSIONS Psychotropic medications are commonly prescribed to Medicaid beneficiaries. An educational intervention consisting of only prescriber mailings is likely not sufficient to improve quality of care or reduce drug costs.