We previously demonstrated that the estrogenicity of either bisphenol A [BPA; 2,2-bis(4-hydroxyphenyl)propane] or bisphenol B [BPB; 2,2-bis(4-hydroxyphenyl)butane] was increased several times after incubation with rat liver S9 fraction (Yoshihara et al., 2001). This metabolic activation, requiring both microsomal and cytosolic fractions, was observed with not only rat liver, but also human, monkey, and mouse liver S9 fractions. To characterize the active metabolites of BPA and BPB, we investigated the structures of the isolated active metabolites by negative mode LC/MS/MS and GC/MS. The active metabolite of BPA gave a negative mass peak at [M-H]− 267 on LC/MS and a single daughter ion at m/z 133 on MS/MS analysis, suggesting an isopropenylphenol dimer structure. Finally, this active metabolite was confirmed to be identical with authentic 4-methyl-2,4-bis(p-hydroxyphenyl)pent-1-ene (MBP) by means of various instrumental analyses. The corresponding peaks of the BPB metabolite were [M-H]− 295 and m/z 147, respectively, suggesting an isobutenylphenol dimer structure. Further, coincubation of BPA and BPB with rat liver S9 afforded an additional active metabolite(s), which gave a negative mass peak at [M-H]− 281 and two daughter ion peaks at m/z 133 and m/z 147 on MS/MS analysis. These results strongly suggest that the active metabolite of either BPA or BPB might be formed by recombination of a radical fragment, a one-electron oxidation product of carbon-phenyl bond cleavage. It is noteworthy that the estrogenic activity of MBP, the active metabolite of BPA, is much more potent than that of the parent BPA in several assays, including two reporter assays using a recombinant yeast expressing human estrogen receptor α and an MCF-7-transfected firefly luciferase plasmid.