OBJECTIVE: To determine the efficacy, tolerability, and safety of duloxetine when added to oral nonsteroidal anti-inflammatory drugs (NSAIDs) in patients with osteoarthritis (OA) of the knee with pain of moderate or greater severity. RESEARCH DESIGN AND METHODS: This was a 10-week randomized, double-blind, flexible-dose (duloxetine 60/120 mg/day), placebo-controlled trial that enrolled adult outpatients who had persistent moderate pain (≥4 on a 0-10 numerical rating scale) due to OA of the knee, despite, per protocol, having received optimized oral NSAID therapy (specific drug, dose, and frequency at investigator discretion). CLINICAL TRIALS REGISTRATION: ClinicalTrial.gov identifier: NCT01018680. MAIN OUTCOME MEASURE: Patients entered daily pain ratings in a telephone-based diary. The primary efficacy outcome was the weekly mean of the daily average pain rating at week 8. Safety outcomes were assessed during the entire 10-week study. RESULTS: A total of 524 patients randomly received duloxetine 60/120 mg/day (N = 264) or placebo (N = 260). In total, 74% of the patients completed the study. Mean age was 61 years (SD 9.2), 57% were female, and 81% were white. Duloxetine-treated patients had significantly greater pain reduction at week 8 (p < 0.001) than placebo-treated patients. In addition, relative to placebo at week 8, duloxetine-treated patients had significant improvements in physical function as measured by the Western Ontario and McMaster Universities Osteoarthritis Index (p < 0.001), and Patient Global Impression of Improvement (p < 0.001). Compared to placebo, significantly more nausea, dry mouth, constipation, fatigue and decreased appetite were reported by patients taking duloxetine (each p < 0.05). Discontinuation due to adverse events occurred more commonly in the duloxetine group than the placebo group (p = 0.03). CONCLUSION: Duloxetine added to oral NSAID therapy provided additional significant pain reduction, improved function, and patient-rated impression of improvement. Adverse events were consistent with those seen in previous duloxetine trials. The short duration of the study may not reflect the longer term efficacy and safety of NSAID/duloxetine cotherapy.