Polychlorinated biphenyls (PCBs) with hydrophobicity, lipophilicity and high toxicity, are a group of synthetic persistent organic contaminants, and have caused people widespread concern. In this work, density functional theory (DFT) was employed to calculate some structural parameters of PCBs, such as the negative charge density of the benzene ring (Q), E-LUMO, electrophilicity index (omega), and the relationship of toxicity of PCBs with coplanarity, the number of substituted chlorines (N-Cl), Q, and omega was also investigated. Using SPSS17 program, the relevancy of these structural parameters with n-octanol-water partition coefficients (K-OW) was analyzed, and multiple linear regression equations of 1g K-OW for PCBs were constructed. Tests of these quantitative structure-property relationship (QSPR) equations were performed to ensure the stability using cross-validation method, and those equations were also used to predict the toxicity of PCBs. The established QSPR equation based on N-Cl and omega is simple and reliable, and the predicted 1g K-OW values of PCBs agree well with experimental results.