Although common, hand osteoarthritis is controversial and rarely used as a model for clinical trials in osteoarthritis. We found only 13 therapeutic trials conducted in digital or trapeziometacarpal osteoarthritis between 1983 and 1994. Eleven of these trials were published. Seven were on nonsteroidal antiinflammatory drugs given either per os (two trials, meclofenamate and ibuprofen) or percutaneously (one trial each on etofenamate, ibuprofen, and ketoprofen gel, and two trials on niflumic acid gel), three were on symptomatic slow-acting drugs (glycosaminoglycanes in two trials and chondroitin sulfate in one), and three were on miscellaneous agents (the muscle relaxant idrocilamide, as a gel; the antisubstance P agent capsaicin, also as a gel; and a spa treatment). We have reviewed the methodology and findings of these trials with the goal of determining the optimal approach to realize better standardized trials in the next future for identifying symptomatic slow-acting drugs and/or "chondroprotective" agents with beneficial effects in digital osteoarthritis.