Following several years of development and validation, we applied the WOMAC Osteoarthritis Index as the principal outcome measure in a double blind randomized parallel trial of Meclomen (100 mg po tid) and Voltaren (25 mg po tid). Statistically significant improvements in clinical status were noted in both treatment groups. At the doses studied, between drug differences favoring Meclomen were observed in pain and stiffness, no difference being noted in physical function. No significant between drug difference was noted in tolerability at these same doses. Our study also demonstrated that the relative efficiency of WOMAC was similar to that of the Lequesne and Doyle indices. Finally, we defined the standard deviation necessary to calculate sample size for future studies using the WOMAC index, both for studies based on static scores and those based on change scores.