In familial adenomatous polyposis, sulindac-induced polyp regression has been reported by several authors. In this study, the goal was to confirm these results by a randomized, placebo-controlled, double-blind crossover study in 10 patients with rectal polyps that had been previously treated by colectomy and ileorectal anastomosis. Patients received sulindac, 300 mg/day, or placebo during two 4-month periods separated by a 1-month wash-out phase. One patient was not compliant and was excluded. With sulindac, the authors observed a complete (6 patients) or almost complete (3 patients) regression of the polyps. With placebo, the authors observed an increase (5 patients), no change (2 patients), and a relative decrease (2 patients) in the number of polyps. The difference between sulindac and placebo was statistically significant (P less than 0.01). In biopsy specimens of polyps and normal rectal mucosa of 6 patients, the authors conducted an immunohistochemical study of the cellular proliferation index using the Ki 67 monoclonal antibody (Ki 67 index), at the beginning and at the end of each treatment period. They were not able to show a sulindac-induced modification of the Ki 67 index. The authors conclude that sulindac is effective in inducing the regression of rectal polyps in familial adenomatous polyposis.