Epidemiological evidence from different countries worldwide has suggested a positive association between the availability of fat in the diet and variations in breast cancer mortality rate. A voluminous amount of information is also available in the literature linking increased fat consumption, particularly polyunsaturated fat, and stimulation of mammary tumorigenesis in animal models. In the past few years, our laboratory has been studying the impact of several confounding factors that could modulate the enhancing effect of fat on neoplastic development of the mammary gland in female rats which are treated with a carcinogen. It is our conclusion that fat promotes mammary carcinogenesis only under a very stringent set of conditions which might not be duplicated in the arena of fat intake and human breast cancer risk. Previous studies on fat and mammary cancer in experimental models have used young virgin rats which are given a dose of carcinogen at a particular age. The question arises as to whether the promoting effect of fat might be a consequence of the characteristics of the model. We have supportive evidence showing that the following criteria must be satisfied in order for fat enhancement of mammary carcinogenesis to be manifested: (a) carcinogen administered at a time when the mammary gland is exquisitely susceptible to tumor induction, (b) animals maintained on a semipurified diet, (c) ad libitum feeding necessary, and (d) unusually high requirement of linoleic acid for tumor development. On the other hand, the stimulatory effect of fat is attenuated or sometimes even negated by (a) feeding of a natural ingredient diet, (b) submaximal calorie intake, and (c) previous history of pregnancy and lactation. Given the spectrum of confounders that are inherent in epidemiological studies linking fat intake and breast cancer, including differences in lifestyle, reproductive history, eating habits, as well as complexity of the total diet, our findings suggest that there may be a need to reevaluate the validity of extrapolating animal data that are obtained under a highly defined set of conditions to the etiological significance of dietary fat in human breast cancer.