OBJECTIVES:  The purpose of this study was to compare adherence and persistence for patients treated with duloxetine across several diagnoses categories: major depressive disorder (MDD), fibromyalgia (FM), osteoarthritis (OA), and chronic lower back pain (CLBP).  METHODS: Patients age 18-64 initiating duloxetine treatment during 2008 were identified in the Thomson Reuters MarketScan® Database. The index event was defined as the first duloxetine prescription filled during the study period with no duloxetine coverage during the previous 90 days. Patients were assigned to disease-category cohorts on the basis of ICD-9 codes recorded on medical claims dated within +1/- 3 month of the initial duloxetine prescription. Adherence was measured over both 365- and 90-day post duloxetine initiation periods as the percent of patients with a medication possession ratio ≥ 0.8. Persistence was measured over a 365-day post-initiation period as the percent of patients with length of therapy ≥ 180 days.  χ2-tests were used to compare differences in adherence and persistence across patient cohort.   RESULTS: A total of 18,406 patients with one of the 4 identified diagnosis categories were identified as initiating duloxetine in 2008: MDD (8,334), FM (3,630), OA (1,458) or CLBP (4,984).  Adherence was directionally greater among patients with MDD (37.3%) versus FM (35.3%) or OA (35.7%), and statistically significantly greater than CLBP (29.9%; p<0.005).  Comparisons of 90-day adherences were similar, with MDD (58.8%) directionally higher than FM (54.1%) or OA (57.8%), and statistically significantly greater than CLBP (50.0%; p<0.005). Comparisons of persistence were similar to adherence. For example, persistence was 47.5% for MDD versus 38.7% for CLBP (p<0.005).   CONCLUSIONS: Duloxetine adherence and persistence were similar among patients with MDD, FM, and OA, and significantly less among those with CLBP.  These relationships were consistent across variations in technical assumptions employed in calculating the study measures.