We compared the effects of identical amounts but different proportions of dietary n-3 polyunsaturated fatty acids (PUFAs) on N-methyl-N-nitrosourea (MNU)-induced mammary cancer in a rat model. The ability of dietary docosahexaenoic acid (DHA) and eicosapentaenoic acid (EPA) to suppress mammary cancer was evaluated. Female Sprague-Dawley rats were randomly assigned to three groups and maintained on diets containing 10% fatty acid consisting of EPA, a 1:1 mixture of EPA-plus-DHA, or DHA. The experimental diet was started after administration of MNU at 49 days of age, and the rats were maintained on the respective diets until the largest mammary tumor reached >1 cm in diameter or until the end of the study period (20 wk after MNU). All histologically detected mammary carcinomas were evaluated, irrespective of size. The DHA diet was associated with significant suppression of the carcinogenic effect of MNU compared with the EPA and EPA-plus-DHA diets: tumor incidence decreased to 23% (3/13) compared with 73% (11/15) and 65% (12/17) (P < 0.01 and P < 0.05, respectively); tumor multiplicity decreased to 0.23 compared with 1.67 and 1.59 (P < 0.01 and P < 0.05, respectively). There was no significant difference in tumor latency among the DHA, EPA, and EPA-plus-DHA groups (119, 105, and 117 days, respectively). Over 20 wk, the fatty acid composition of serum and mammary fat tissue reflected differences in the dietary n-3 PUFAs. Although DHA suppressed MNU-induced mammary carcinogenesis more effectively than EPA, generalized steatosis including mammary fat tissue appeared in all three groups.