To evaluate the efficacy of an oral nonsteroidal anti-inflammatory drug in the treatment of lateral epicondylitis. Multicenter double-blind randomized controlled trial in which the following hypothesis was tested: whether diclofenac sodium provided a 20% or greater improvement over rest and cast immobilization in the response rate to treatment of lateral epicondylitis beyond and over rest in an experimental group compared with a control group after 4 weeks of treatment. Recruitment from urban general practices and referrals to 4 university hospitals. During a 1-year period, 206 subjects aged 18 to 60 years with lateral epicondylitis were recruited from the clientele treated by family physicians. Thirty subjects refused to participate and 47 presented with exclusion criteria, leaving 129 subjects who entered the study. One subject withdrew after 21 days. The experimental group was treated with a daily dose of diclofenac sodium (150 mg) for 28 days, while the control group received a placebo during the same period. In addition, both groups were immobilized in a cast for 14 days and were told not to perform repetitive movements of the involved limb for 21 days. Measuring instruments consisted of grip strength measurements with a squeeze dynamometer, a visual analog pain scale, a visual analog function scale, and an 8-item pain-free function index. A statistically and clinically significant reduction of pain was associated with treatment with diclofenac, but no clinically significant difference in grip strength or functional improvement could be detected between the 2 groups. Secondary effects (diarrhea and abdominal pain) were significantly more frequent in the diclofenac-treated group. Taking into account the limited improvement noted over rest and cast immobilization and the number of associated adverse events, it is difficult to recommend the use of diclofenac in the treatment of lateral epicondylitis at the dosage used in this study.