We have attempted to elucidate the "natural history" of ulcerative colitis by studying the placebo groups of 11 controlled trials of 185 patients with active disease. Most (182) of the patients were mild to moderately ill, but 3 had severe colitis. They were studied by endoscopic, pathologic, and/or clinical criteria and followed for 15-42 days. Up to 52% improved clinically and 59% sigmoidoscopically. Relapses commonly occurred within 2 months of these "spontaneous" improvements. In another six trials of 174 patients in remission receiving only placebo therapy (134 followed for 6 months and 40 for 12 months), up to 51% remained in remission. Favorable predictive factors for continued remission included a normal rectal mucosa on sigmoidoscopy, limited extent of colitis, and remission recently achieved by active therapy or maintained by therapy for at least 1 year before study entry. Thus, patients can get and stay better with no "specific" therapy. We have clarified advantageous predictive factors.