Torula yeast diets deficient (less than or equal to 0.02 ppm) in selenium (Se) and a control diet supplemented with sodium selenite (0.1 ppm Se), were fed to 52 male Sprague-Dawley rats per diet for 23 wk following weaning. 1,2-dimethylhydrazine (DMH) was administered (20 mg/kg body weight) in 20 weekly i.p. injections after 3 wk of feeding. After 23 wk, 67% of the rats fed the Se-deficient diet had intestinal adenocarcinomas versus 63% on the 0.1 ppm Se diets. Diet also had no effect on the number or size of tumors per tumor-bearing animal or on the location of the tumors within the colon. The effects of Se deficiency on the hematological variables of white blood cell count, hematocrit, serum urea nitrogen and cholesterol concentrations were examined. Serum urea nitrogen levels were lower in Se-deficient DMH-treated rats (9.6 +/- 0.7 vs. 13.7 +/- 0.9 mg/dl, P less than 0.01) and serum cholesterol was higher in Se-deficient DMH-treated animals (69.0 +/- 5.5 vs. 50.7 +/- 3.9 mg/dl, P less than 0.05). Body weights of the Se-depleted group were lower in the DMH-treated animals (P less than 0.01) although food consumption did not differ. Controls without DMH did not show differences due to Se status for any variable examined. Se deficiency appears to affect DMH toxicity but nutritional adequacy (0.1 ppm Se) does not inhibit tumor development. The results of this study do not support the belief that Se deficiency enhances colon carcinogenesis.