Fish oil has been shown to have a protective effect in some cancer models. To determine whether fish oil alters skin tumorigenesis, a study was designed using the initiation-promotion mouse skin carcinogenesis model, feeding mice during the promotion stage a constant overall amount of dietary fat (10%) in which the levels of menhaden oil (MO) varied from 0 to 8.5% or corn oil (CO) at 10%. SENCAR mice were initiated with 10 nmol dimethylbenz[a]anthracene. Two weeks later mice were divided into five groups and maintained on one of the following AIN-76 based diets consisting of: 8.5% coconut oil (CT)/1.5% CO (diet A); 1% MO/7.5% CT/1.5% CO (diet B); 4% MO/4.5% CT/1.5% CO (diet C); 8.5% MO/1.5% CO (diet D); or 10% CO (diet E). Two weeks later, promotion with twice weekly applications of 1 micrograms 12-O-tetradecanoylphorbol-13-acetate (TPA) was begun and continued for 24 weeks. No statistically significant differences in kcal food consumed or body wts were observed between diet groups during the study. The final papilloma and carcinoma incidence was not different among the diet groups. However, differences were seen in the rate of papilloma appearance with the group fed diet E (10% CO) being the slowest and diet B being the most rapid. In a parallel study, ornithine decarboxylase activity, a suggested marker of promotion, was greatly elevated in the epidermis of all TPA-treated mice and the effect of diet tended to reflect the different rates of tumor formation observed among the groups. These data indicate that the diets containing fish oil were not protective in the final incidence of tumor formation and suggest that a better understanding of the complex interactions is warranted before recommendations are made to alter the human diet for cancer prevention.