Three quantitative structure-retention relationship (QSRR) models have been developed for gas chromatographic relative retention indices (RIs) of polychlorinated naphthalene (PCN) congeners. With computation at HF/6-31G* and B3LYP/6-31G* levels in Gaussian98 program, the structural parameters of PCNs were obtained and used as chemical descriptors to correlate with their gas chromatographic RIs by multiple linear regression analysis, resulting in model 1 and model 2. The squared correlation coefficients (r 2 ) of the first two models are 0.9814 and 0.9957, and the crossvalidation correlation coefficients (q 2 ) are 0.9771 and 0.9948, respectively. On the other hand, the third model, model 3, was developed using the relative position of chlorine substitution as molecular descriptor, and its multiple squared correlation coefficient is r 2 ¼ 0.9967 and q 2 ¼ 0.9959, suggesting model 3 is practical and especially has a optimum predictive power.