Male F344 rats, 9 weeks of age, were given multiple intrarectal administrations of N-methyl-N-nitrosourea (NMU) at 0.5 mg/dose twice a week for a total of 8, 12, or 16 doses. Five days after the final NMU instillation, rats were placed on one of four diets; chow with retinoid vehicle, chow with 654 mg N-ethylretinamide per kg diet, chow with 686 mg N-(2-hydroxethyl)retinamide per kg diet, or 406 mg retinylidene dimedone per kg diet. Groups of 40--50 rats receiving 16, 12, or 8 total doses were sacrificed 32, 44, or 52 weeks after the initial NMU dose, respectively. The number of rats with colon tumors and the number of tumors per tumor bearing rat in each dosage regimen was compared to appropriate controls. Organ weights for testis, liver, and kidneys were similar for controls and retinoid treated animals and no histopathologic lesions indicative of retinoid toxicity were found. Unusual proliferative lesions of the colon in rats of all dosed groups included hyperplastic and inflammatory polyps and sarcomas. The feeding of diets containing these three retinoids did not significantly modify the incidence or multiplicity of colon tumors observed under these conditions.