OBJECTIVES: Pharmacy databases can yield important information about drug utilization and costs. This study sought to examine changes in prescription utilization and costs among atypical antipsychotic (AAP) subjects initiating lurasidone or aripiprazole therapy.  METHODS: Adults filling lurasidone or aripiprazole prescriptions from 2/3/2011—6/30/2013 were identified in the Walgreens pharmacy-database.  Treatment-naive monotherapy subjects (no AAP prescriptions before-and-after index prescription) with ≥12-months pre-/post-index continuous enrollment were eligible. Lurasidone subjects were compared to a 1:1 matched random sample of aripiprazole subjects.  Baseline demographics and health-insurance status were compared between cohorts. Mental-health prescriptions (antianxiety agents, antidepressants, antipsychotics, psychotherapeutic, and neurologic agents) were identified using Generic Product Identifier for the National Drug Code numbers on pharmacy claims. Differences in mean changes (post-pre) in all-cause prescription fills all-cause prescription costs, mental-health prescription fills and mental-health costs were compared using t-tests.   RESULTS: Each cohort included 4,595 subjects (lurasidone vs aripiprazole: 69.3% vs 72.2% female, mean ages 41.0 vs 43.4 years). Most subjects were commercially-insured (39.9% vs 48.6%), followed by State-Medicaid (24.4% vs 18.7%), Medicare-Part-D (22.8% vs 18.7%), and Managed-Medicaid (6.8% vs 5.6%). Lurasidone subjects had lower 30-day equivalent co-pays ($42.02 vs $56.63). Subjects were more likely to be prescribed lurasidone  by psychiatrists (78.0% vs 57.3%) and less likely by general-practitioners (3.1% vs 23.0%). Overall, lower mean increases in all-cause prescription fills (11.3 vs 12.3;p=.09) and mental-health prescription fills (7.2 vs 8.0;p<.01) were observed for lurasidone than aripiprazole subjects. Additionally, mean differences in all-cause prescription costs ($2,388 vs $3,080;p<.01) and mental-health prescription costs ($2,123 vs $2,810;p<.01) were lower for lurasidone than aripiprazole subjects. Similar patterns of mean changes in prescription utilization and costs were found in commercial, Medicaid and Medicare subjects. CONCLUSIONS: In this national-US pharmacy-database analysis comparing subjects initiating branded AAPs, lurasidone subjects had fewer mean changes in all-cause and mental-health prescriptions and lower mean increases in associated costs than aripiprazole subjects.