The development of novel ionophores for ion-selective sensors is a time-consuming and tedious process requiring synthesis of candidate substances, preparation of plasticized polymeric membranes, and their thorough characterization with traditional protocols to assess sensitivity, selectivity and detection limits for target ions. The vast amount of literature data accumulated on various ion-selective sensors allows for significant facilitation of the development through in silico experiments. In this report, we performed the feasibility study on the prediction of potentiometric Mg2+/Ca2+ selectivity for various amide ligands using quantitative structure-property relationship (QSPR) modeling. The approach proved to be promising for ionophore screening purposes with achieved precision in prediction of the selectivity coefficient logK(Mg2+/Ca2+) of 0.5 in the range from -1.7 to +2.3. The study also shows a route for prediction of new potential ionophores with high selectivity values.