The molecular structures of 117 nitrogen-containing polycyclic aromatic compounds (N-PACs) were described by a method of molecular structural characterization (MSC) called molecular electronegativity interaction vector (MEIV). The samples were divided into a training set and a test set. For the training set, a quantitative structure-retention relationship (QSRR) model was built up by multiple linear regression (MLR) and the model was evaluated by performing the cross validation with the leave-one-out (LOO) procedure. The correlation coefficient (R) and the cross-verification correlation coefficient (R CV ) of the model were 0.992 and 0.991, respectively. Moreover, the model was evaluated by the test set and satisfactory results with a correlation coefficient (R test ) of 0.993 were obtained. The results suggested good stability and predictability of the model.