The effect of the combination of sulfasalazine and prednisone has been compared with that of prednisone and placebo in 89 actively symptomatic patients with Crohn's disease in a double-blind, randomized, multicenter controlled trial. The combination was less effective than prednisone alone in treatment of active symptomatic disease. The probability of obtaining this result, if sulfasalazine truly has a clinically useful effect equal to or greater than that specified in the calculation, is less than 1%. Patients who were in remission at the end of 8 wk were rerandomized to receive either the two drugs together or prednisone plus placebo while repeated systematic attempts to withdraw prednisone were made over the next 6 mo. Sulfasalazine showed no prednisone-sparing effect as judged either by outcome ranking or total dose of prednisone consmed by the two treatment groups. However, in this comparison the probability is greater than 5% that, given the results observed, a clinically useful effect of sulfasalazine of specified minimum degree truly exists. It was possible to withdraw prednisone from 25% of patients at the first attempt and ultimately in 37%.