Objective: To assess the rate of substance abuse among insomnia patients prescribed nonbenzodiazepine receptor agonists (nBZRAs), benzodiazepines (BZDs), and melatonin receptor agonists (MRAs). Methods: This analysis used a retrospective matched cohort design with data from Florida Medicaid recipients 18 to 64 years old. Patients initiating nBZRA, BZD, or MRA therapy between July 1, 2002 and March 31, 2006, without a history of substance abuse, were included in this analysis. One MRA patient was matched to three nBZRA and three BZD patients based n demographic and clinical (history of insomnia, depression, anxiety) characteristics. Service dates and days of medication supply variables from pharmacy claims established periods of treatment exposure, while substance abuse related to sedatives and hypnotics was identified using International Classification of Diseases 9th edition Clinical Modification (ICD-9-CM) codes for the same time frame. Rates of substance abuse (SA) were calculated as the number of patients with a claim for SA per 100 patient-years of treatment exposure. Results: A total of 140 MRA patients were identified and matched to 420 nBZRA and 420 BZD patients by age (mean 46 years), sex (64% female), race (57% white), depression (50%), and anxiety (14%). The mean Charlson co-morbidity score was higher for MRA patients (1.4) than nBZRA or BZD patients (0.6 and 0.7, respectively, P<0.01). The SA rate per 100 patient-years of exposure was 1.84 and 2.51 for nBZRA and BZD patients, respectively. There was no evidence of substance abuse among the MRA cohort. Conclusion: These initial data suggest that MRA patients were less likely to use medical services for substance abuse than nBZRA or BZD patients. However, given the small number of patients in this study, no definitive conclusions can be drawn. Additional data will be needed to confirm these findings.