OBJECTIVES: This study investigated the effect of Medicare Part D on prescribing patterns and drug utilization of non-benzodiazepine sedative hypnotics. METHODS: Time-series analyses were conducted using data from National Ambulatory Medical Care Survey (NAMCS).  Subjects were derived from US. outpatient visits between 2002 and 2009 where the primary payment source was Medicare and at least one non-benzodiazepine sedative hypnotic drug was prescribed.   Data trends were graphically plotted and further analyzed using segmented regression to estimate the effects of the Medicare Part D on drug utilization.   A weighted multivariate logistic regression was conducted to predict the maximum likelihood of prescribing pattern associated with patient and physician socioeconomic characteristics.  All analyses utilized SAS PROC SURVEY applications to adjust for the complex sampling design employed by NAMCS database. RESULTS: An estimated 31.52 million of Medicare beneficiaries received at least one non-benzodiazepine prescription between 2002 and 2009 during their outpatient visits.  After Medicare part D in 2006, there was a notable increase (24%) in Medicare outpatient visits between 2006 and 2009. In the same time period, prescribing of non-benzodiazepine sedatives increased significantly by 46.3%.  The results from segmented regression indicate that the implementation of Medicare Part D drug benefits has significantly increased the sedative utilization in Medicare population (P=0.0001).  Multivariate logistic regression revealed that patient gender, geography, chronic condition, and physician specialty all play an important role in determining the utilization pattern of non-benzodiazepine sedatives.  CONCLUSIONS: Our study indicated that the use of non-benzodiazepine hypnotics increased dramatically after Medicare Part D.  Increased utilization may also be related to the switching effect from benzodiazepine formulary exclusion and/or antidepressant off-label use for insomnia pharmacotherapy.  These findings show the importance of using data analysis to identify substantial consequences from policy implementation and the need to provide additional guidance to insurers on how to effectively monitor prescribing patterns.