In contrast to 5-methylchrysene and 5,9-dimethylchrysene, 5,6-dimethylchrysene and 5,7- dimethylchrysene are weak tumor initiators on mouse skin. In order to investigate the basis for this, we have evaluated the mutagenic activities toward Salmonella typhimurium TA 100 and reactivity with DNA of (±)-anti-1,2-dihydroxy-3,4-epoxy-1,2,3,4-tetrahydro-5,6-dimethylchrysene (anti-5,6-diMeC-1,2-diol 3,4-epoxide) and anti-5,7- and anti-5,9-diMeC-1,2-diol 3,4-epoxide. The tumorigenic activities of anti-5,6- and anti-5,7-diMeC-1,2-diol 3,4-epoxides in newborn mice were also investigated anti-5,9-diMeC-1,2-diol 3,4-epoxide was the most mutagenic of the three diol epoxides, anti-5,6-diMeC-1,2-diol 3,4-epoxide was highly tumorigenic in newborn mouse lung, with activity significantly greater than that of either anti-5-MeC- or anti-5,7-diMeC-1,2-diol 3,4-epoxide. Although the amounts of total binding of the diol epoxides to calf thymus DNA were similar, anti-5,6-diMeC-1,2-diol 3,4-epoxide bound extensively to deoxyadenosine residues. High binding to deoxyadenosine is related to the presence of a sterically hindered bay or fjord region as present in 5,6-diMeC, 7,12-dimethylbenz[a]anthracene, benzoyl[g]chrysene, and benzo[c] phenanthrene. The conformations of the anti- and syn-diol epoxides of 5,6-diMeC and benzo[c] phenanthrene were similar, with both having pseudodiequatorial hydroxyl groups, in contrast to less sterically crowded diol epoxides. The high tumorigenicity of anti-5,6-diMeC-1,2-diol 3,4-epoxide in newborn mice is of interest with respect to its high deoxyadenosine binding. However, a relationship between deoxyadenosine binding and tumorigenicity among other chrysene-1,2-diol 3,4-epoxides was not apparent. The results of this study provide further insight into the structural requirements favoring diol epoxide reactivity with deoxyadenosine in DNA and tumorigenicity in newborn mice.