In an attempt to determine the requirement of essential fatty acid for dimethylbenz(a)anthracene-induced mammary tumorigenesis, rats were fed diets containing different levels of linoleate: 0.5, 1.1, 1.7, 2.2, 3.5, 4.4, 8.5, or 11.5%. Each diet contained 20% of fat by weight, with varying amounts of coconut oil and corn oil added to achieve the desired levels of linoleate. Mammary tumorigenesis was very sensitive to linoleate intake and increased proportionately in the range of 0.5 to 4.4% of dietary linoleate. Regression analysis indicated that a breakpoint occurred at 4.4%, beyond which there was a very poor linear relationship, suggesting the possibility of a plateau. From the intersection of the regression lines in both the upper and lower ranges, the level of linoleate required to elicit the maximal tumorigenic response was estimated to be around 4%. The differences in tumor yield could not be correlated with changes in prostaglandin E concentration in the mammary fat pads of normal animals maintained on similar diets, suggesting that linoleate may act by some other mechanism to stimulate mammary tumorigenesis.