There is a considerable variation in the diets used in studies on the influence of dietary fat on rat mammary cancer. In view of the fact that diet is the most remarkable factor in these studies, the aim of this work was to define two experimental diets, one of them normal (N3) and another hyperlipidic (HL20), both allowing the normal growth of the rat and neither of them containing factors that could unspecifically affect mammary carcinogenesis. Semisynthetic diets were selected instead of natural ones. A normal diet (3% corn oil, 18% casein, 67.9% dextrose) and a hyperlipidic diet (20% corn oil, 23% casein, 45.9% dextrose) were defined for the rat. Both diets also contain 5% cellulose, 5.9% salt mix and 0.24% vitamin mix. In order to avoid the influence of the above mentioned unspecific factors, the control of specificity and quality of nutrients is proposed as an essential measure. Moreover, it is also necessary to adopt measures to avoid the presence of fatty acid metabolites, including the calculation of the necessary vitamin E, selenium and sulfur amino acid and the determination of factors potentially able to stimulate or inhibit carcinogenesis such as phenolic antioxidants, retinoids or the trans isomer of fatty acids. On the other had, casein, dextrose, choline and folic acid contents were modified in order to equilibrate the lipid increase experimentally introduced in the HL20 diet or to ensure the normal maintenance of the animals' metabolism. The method used is based on the concept of quality assurance.(ABSTRACT TRUNCATED AT 250 WORDS)