OBJECTIVES: Safety studies have shown that risks associated with use of non-steroidal anti-inflammatory drugs (NSAIDs) are related to dose and release form; however, there is little US evidence. This study assessed the relationships between diclofenac dose and release form on gastrointestinal (GI), cardiovascular (CV), and renal events using US health care claims. METHODS: The MarketScan^© Commercial Claims and Encounters database (2008-12) was used to analyze the risks of GI (upper GI bleed/perforation [UGIB], lower GI bleed [LGIB]), CV (myocardial infarction [MI], stroke, and congestive heart failure [CHF]), and renal events by diclofenac dose category relative to no current NSAID use and extended/delayed release (ER) relative to immediate release (IR) form among new adult users of diclofenac. Patients with prior history of GI, CV or renal disease were excluded. Hazard ratios (HRs) were estimated using multivariable Cox regression models with total daily dose (TDD) and release form as time-dependent covariates. RESULTS: In total, 851,549 diclofenac users (57% female; median age 50 years) met the initial study inclusion criteria. At diclofenac initiation, median TDD was 150mg (range: 1-400mg) and most (88.25%) prescriptions were ER. The HRs (95% CIs) for TDD >150mg (vs no current exposure) were: UGIB: 3.48 (2.25-5.37) for ER and 2.59 (1.55-4.33) for IR; LGIB: 1.58 (1.31-1.91) for ER and 1.37 (1.11-1.70) for IR; MI: 1.10 (0.74-1.62) for ER and 1.02 (0.66-1.60) for IR; stroke: 1.08 (0.90-1.31) for ER and 1.09 (0.88-1.34 for IR); CHF: 1.39 (1.14-1.70) for ER and 1.38 (1.10-1.74) for IR; and renal failure: 2.25 (1.84-2.76) for ER and 2.05 (1.60-2.61) for IR. CONCLUSIONS: In an analysis of US health care claims, increased risks of certain adverse events were associated with higher doses compared with no current use and with ER compared with IR form among new users of diclofenac with no prior history of the event.