Breast cancer may be the consequence of free radical damage, which is partially caused by the excessive intake of dietary fat and imbalances in antioxidant scavenger system;. In this experiment, we examined! the effects of dietary peroxidizability index (PI) values on hepatic thiobmbituric acid reaction substances (TBARS) and antioxidant enzyme activities in rats treated with 7,12-dimethylbenz[$alpha$]anthracene (DMBA). Female Sprague-Dawley rats were used and 7,12-DMBA (20 mg/kg body weight) was gastrically intubated at seven weeks of age in order to induce mammary tumors (MT). The levels of dietary PI were 36, 81, 126 and 217 (LPI, MLPI, MHPI and HPI), while dietary polyunsaturated/saturated fatty acids ratio was maintained at the same level (1.0). Fat used in the experiment was mixed with soybean oil, com oil, palm oil, perilla oil, sesame oil, fish oil, and beef tallow. Experimental diets were given for the following 20 weeks. We measured tumor numbers and weights, and then assayed the hepatic TBARS levels and antioxidant enzyme activities such as superoxide dismutase (SOD), catalase, glutathione peroxidase, glutathione-S-transferase (GST) and glutathione reductase (GR). The incidence of Mr was the lowest in the MHPI group. The hepatic TBARS level was significantly raised with increasing dietary PI value. The hepatic SOD and GR activities were differed significantly by dietary PI value. The hepatic SOD activity was negatively correlated with dietary PI value and GR activity was the highest in the rats fed the MHPI diet. When the dietary P/S ratio is kept at 1.0, adequate dietary PI value (PI value of 126) may reduce the incidence and growth of Mr, but this benefit may be lost with higher dietary PI value. These results suggest that the awareness of dietary PI values may help to decrease breast cancer incidence and growth.