OBJECTIVES: To assess the cost-effectiveness of long-acting risperidone (LAI-RIS), oral risperidone (RIS), olanzapine (OLA), quetiapine (QUE), ziprasidone (ZIP), aripiprazole (ARI), and haloperidol decanoate (HAL-DEC) in patients with schizophrenia over one-year from a health care system perspective. METHODS: Published medical literature, an unpublished consumer health database, and a clinical expert panel were utilized to populate a decision tree model. The model captured rates of compliance and relapse, frequency and duration of relapse, adverse events, resource utilization and unit costs. Outcomes included percentage, number and duration of relapses per patient per year and direct medical costs. RESULTS: The mean days of relapse requiring hospitalization per patient per year were 28 HAL-DEC, 18 RIS, OLA, QUE, ZIP and ARI, 11 LAI-RIS, while the mean days of exacerbation not requiring hospitalization were eight HAL-DEC, five RIS, OLA, QUE, ZIP and ARI, three LAI-RIS. Direct medical cost savings with LAI-RIS compared to RIS, OLA, QUE, ZIP, ARI, and HAL-DEC were $161, $1425, $508, $259, $1068, and $8224, respectively. CONCLUSIONS: Long-acting risperidone may lead to substantially lower rates and fewer days of symptom exacerbation and hospitalization compared to currently available treatments. These lower rates translate into direct medical cost savings with the use of long-acting risperidone.