A quantitative structure-retention relationship (QSRR) technique is a powerful approach which has been widely used to predict the retention time of analytes in numerous areas of chromatography. Current trial-and-error techniques for chromatographic method development, such as the selection of optimal chromatographic conditions, are tedious, time-consuming and expensive. On the other hand, the selection of chromatographic conditions can be simplified by the prediction of retention time for a given analyte, based on the characteristics of the analyte. In this work, a QSRR method is introduced to model the retention time of inorganic anions in ion-exchange systems based on the chemical structure of the ion, with a view to providing a guide as to the best column for the separation of the target ion. This approach is performed using a database of retention data of inorganic anions forming part of the “Virtual Column” software marketed by Thermo Fisher Scientific. Molecular descriptors for these analytes are generated from Dragon software. A principal component analysis on a large set of molecular descriptors is used to extract an optimal subset which best describes the relationship between the descriptors and retention for analytes. Finally, the model equation for retention time of analytes is derived by multilinear regression (MLR) using the optimal subset of molecular descriptors. Thus, the most suitable column for the separation of target analytes in ion-exchange chromatography can be selected without experimentation, which leads to a reduction of costs and time in chromatographic method development.