OBJECTIVES:  A major concern regarding prior authorization (PA) policies for atypical antipsychotics (AAs) is that they produce undue burden on an already vulnerable population which may worsen outcomes. In October 2008 Ohio Medicaid implemented a PA-policy that contained a unique stipulation exempting psychiatrists. The objective of this study was to determine the impact of this PA-policy on healthcare utilization and cost for patients newly prescribed AAs. METHODS:  Patient-level data analyzed came from Ohio Medicaid fee-for-service claims files June/2007-September/2009. A difference-in-differences regression approach was used to compare outcomes in the PA-active year to the year pre-implementation between patients treated by psychiatrists and those treated by nonpsychiatrists. Patients were stratified based on index-AA prescribing-physician type, and those treated by PA-exempt psychiatrists served as the control. Patients were included if they were ≥18, had an AA claim following 120-days of washout, and had 180 days of post-index healthcare utilization data. Patients with any gap in coverage and those dual-eligible for Medicare were excluded. Logistic regression was used to estimate the policy-attributable effect on all-cause and psychiatric-related hospitalizations, emergency department visits, outpatient visits, and physician office visits. Policy-attributable changes in all-cause and psychiatric-related expenditures were assessed using a generalized linear model with a log-link function and gamma distribution. All regressions included the following covariates to control for confounding: age, sex, race, diagnosis, comorbidities, previous healthcare utilization, index-AA type, and AA-adherence. RESULTS:  1,129 patients from the pre-PA-period (psychiatrist-treated=400; nonpsychiatrist-treated=729) and 2,032 patients from the PA-period (psychiatrist-treated=955; nonpsychiatrist-treated=1,072) were included in the difference-in-differences analysis. No statistically-significant policy-attributable effect was found except for all-cause hospitalizations, which decreased substantially with an OR=0.54 (95%CI: 0.33-0.88, p=0.014). CONCLUSIONS:  Implementation of the PA-policy did not negatively affect patient outcomes, and was shown to reduce the odds of hospitalization. The psychiatrist-exemption implemented by Ohio may represent a PA-design that reduces medication costs while preserving patient quality-of-care.