Many clinicians believe that slow-acting therapeutic agents, such as fold, penicillamine, the antimalarials, and cytotoxic drugs, can retard joint destruction in rheumatoid arthritis. We reviewed 60 published studies employing these drugs to evaluate critically the evidence that drug therapy can slow the radiographic progression of disease. Seventeen studies were found that included radiographic assessment of both treated and control groups; they were analyzed using methodologic criteria known to be important in affecting the results of drug trials. In addition to numerous qualitative methodologic deficiencies, many studies showed inadequacies in sample size and duration of treatment, and the drug dosage used varied from one study to another. We found evidence suggesting that both gold and cyclophosphamide can retard radiographic progression of joint destruction. At present, there are too few technically adequate studies to permit even provisional conclusions concerning other agents.