A double-blind randomized parallel group trial was undertaken to compare the acceptability and efficacy of 2 forms of analgesic treatment, DI-Antalvic (Houde Laboratories, Puteaux, France) (30 mg dextropropoxyphene and 400 mg paracetamol per capsule) and Efferalgan-Codeine (UPSA Laboratories, Rueil Malmaison, France) (30 mg codeine and 500 mg paracetamol per tablet) prescribed for 1 week at doses of 6 capsules/day and 6 tablets/day, respectively, in 141 outpatients with active osteoarthritis of the knee or hip. The principal aim of the trial was concerned with acceptability, with efficacy as its secondary aim. The principal trial criterion was defined as overall assessment of acceptability by the patient at the end of the trial (success or failure) or by treatment dropouts because of an adverse effect (failure). Comparability of the groups was confirmed before any treatment regarding the physical characteristics of the patients, characteristics of osteoarthritis, and the initial level of pain and functional consequences of pain. Results show that the analgesic efficacy of the treatment was similar, but that the acceptability of Efferalgan-Codeine was significantly worse than that of DI-Antalvic: 53% failure with Efferalgan-Codeine versus 29% failure with DI-Antalvic (P = .005). Other trials of the same type would seem necessary (comparison of lower doses, other types of pain) before being able to generally extrapolate such findings.