The c-Jun N-terminal kinases are attractive targets because of their involvement in several diseases. In this work, a combined molecular modeling study for a set of isoquinolones as inhibitors of JNK1 was performed by molecular docking, genetic algorithm–multiple linear regression and comparative molecular field analysis to rationalize the structural requirements responsible for the inhibitory activity of these compounds. Molecular docking study was employed to explore the binding mode of the active compound at the active site of JNK1. Based on the docked conformations, highly predictive 2D, 3D quantitative structure–activity relationship models were developed. The best 2D quantitative structure–activity relationship model was established using genetic algorithm–multiple linear regression method containing four molecular descriptors. The best comparative molecular field analysis model was obtained with a cross-validated coefficient q2 of 0.562, non-cross-validated r2 values of 0.994. The information from quantitative structure–activity relationship models and molecular docking is useful for the design of novel JNK1 inhibitors with improved activities.