N-(2-chloroethyl)-N′-cyclohexyl-N-nitrosoureas　(CCNUs)　are　widely　applied　as　antitumor　agents.　However,　researches　have　indicated　that　their　decomposition　products　are　inclined　to　cause　secondary　cancer.　Hence　it　is　necessary　to　probe　into　both　carcinogenic　and　anticancer　mechanism　of　such　compounds　so　that　effective　anticancer　drugs　with　lower　toxicity　can　be　obtained.　In　the　present　work,　the　DFT-B3LYP　method　with　the　basis　set　of　6-31+G(d,p)　was　employed　to　calculate　the　molecular　geometries　and　electronic　structures　of　17　CCNUs.　The　charges　of　C6　and　(miLogP)2　were　selected　as　quantum　chemical　descriptors.　The　observed　anticancer　activity　along　with　the　two　descriptors　was　used　to　establish　the　quantitative　structure-activity　relationship　(QSARs).　Through　multiple　linear　regression　analysis,　a　model　was　established　as　follows:　log　1/C　=　-0.042　×　(miLogP)2　+　11.783　×　QC6　+　3.237.　Here　C　refers　to　the　molar　concentration　(mol/kg)　of　drugs　producing　a　reduction　of　1000-fold　of　leukemia　cells　to　mice.　This　QSAR　model　can　be　applied　to　predict　the　anticancer　activity　of　CCNUs,　as　well　as　elucidate　their　anticancer　mechanism　reasonably.　©2010　IEEE.