Epidemiological and animal studies have indicated that 17beta-estradiol (E2) is involved in breast cancer; however, the mechanism is unclear. We found that E2 could be activated by epoxidation, resulting in its ability to inhibit nuclear DNA-dependent RNA synthesis, and to bind DNA, forming DNA adducts both in vitro and in vivo. Because epoxidation is required for the activation of many chemical carcinogens, including benzo(a)pyrene, 7,12-dimethylbenz(a)anthracene and aflatoxins, we proposed previously that E2 epoxidation is the underlying mechanism for the initiation of breast cancer. The first part of this review is to present the experimental evidence obtained from this laboratory in support of this hypothesis. Based on these newly discovered insights on E2 epoxidation and its initiation role in breast cancer carcinogenesis, a method to screen chemopreventive agents against breast cancer has been developed. This constitutes the second part of the review. Two examples will be used to illustrate the utility of this screening technique. The effect of fat on breast cancer has been a longstanding but unresolved issue. Epidemiological studies provide conflicting results regarding the association of dietary fat and breast cancer. Because vegetable oils contain various amount of mono- and polyunsaturated fatty acids, they are potential antioxidants. Data are presented to show that commercial vegetable oils, independent of their mono- or polyunsaturated fatty acid content, are all able to prevent the formation of E2 epoxide, as measured by the loss of the ability of E2 to inhibit nuclear RNA synthesis in vitro. Tamoxifen (TAM), an anti-estrogen used for breast cancer treatment, has recently been found to have a strong breast cancer preventive effect. The mechanism for this is unknown. Using the same screening technique, we found that when incubated together with E2 for epoxidation, TAM was able to prevent the formation of E2 epoxide, as evidenced by both the loss of the ability of E2 to inhibit nuclear RNA synthesis and the reduced binding of [3H]-labelled E2 to nuclear DNA in a dose-dependent manner. These experimental results suggest that the breast cancer preventive effect of TAM is to prevent the formation of E2 epoxide through a competitive epoxidation mechanism with E2.