A 3 X 3 factorial experiment was conducted to examine how protein content (8, 16, 32% of kilocalories from casein) and fat content (12, 24, 48% of kilocalories from corn oil) interact to influence 7,12-dimethylbenz(a)anthracene (DMBA)-induced breast carcinogenesis in rats. Forty weanling Sprague-Dawley rats were assigned to each of 9 diets fed ad libitum. After 4 weeks each rat received DMBA (20 mg/kg) via gastric intubation. No substantial statistical interactions of protein and fat were observed on tumor incidence. Increasing dietary corn oil increased the percentage of rats with palpable tumors. Rats fed diets containing 12, 24 and 48% of kilocalories from corn oil showed 35, 49 and 70% tumor prevalence at necropsy, and the total number of tumors per fat level was 65, 81 and 182, respectively. Each doubling of dietary fat concentration approximately doubled the odds of a rat developing a tumor. Multiple tumors were more common with the highest corn oil intake. The odds of finding a second tumor in rats with one tumor increased by a factor of 7.5 when fat kilocalories were increased from 24 to 48% compared to a decrease of one-third when fat kilocalories were increased from 12 to 24%. Dietary corn oil significantly increased the prevalence of adenocarcinomas and adenomas but not fibroadenomas. Dietary protein did not significantly affect tumor prevalence. However, tumors palpated in rats fed 16% of kilocalories as protein regressed more frequently than in rats fed low or high protein diets. Multiple logistic-regression results indicate that, in addition to the response to dietary corn oil, tumorigenesis was increased in rats with greater ad libitum food consumption. This conclusion is supported by reanalysis that used direct rate adjustment and average partial association tests.