Benoxaprofen (600 mg once a day) and indomethacin (25 mg t.i.d.) were compared in a double-blind, randomized, crossover study in 35 ambulatory hospital outpatients suffering from osteoarthritis of the knee. Six patients on each treatment withdrew because of gastric side effects (3 patients withdrew from both treatments) and a total of 5 patients withdrew because of lack of efficacy (2 on benoxaprofen and 3 on indomethacin). Twenty-two patients were suitable for analysis of efficacy and safety. The two drugs were shown to be comparable in efficacy by both the subjective and objective parameters used to assess pain, stiffness, and movement. However, because there was no washout period between administration of the two drugs, there was a tendency for the drug given second (8 benoxaprofen versus 14 indomethacin) to have greater efficacy. Side effects reported were mostly mild in nature and mostly gastrointestinal in origin. A significant reduction in levels of alkaline phosphatase (p less than 0.005) was noted during benoxaprofen therapy.