Electron donors enhance the productivity and isotacticity of products of the Ziegler-Natta catalyzed propylene polymerization. Using the fact that adsorption energies of electron donors to catalyst surface are linearly related to activities, the Quantitative Structure Property Relationship (QSPR) for adsorption energies was performed for a set of 24 compounds from 3 different groups, i.e., phthalates, 1,3-diethers and malonates using the multiple linear regression (MLR) analysis. The QSPR model shows high correlation (R-2 = 0.84, R-cv(2) = 0.83) between adsorption energies and 3 descriptors, i.e., the radius of gyration (50%), the dipole moment (16%), and the forcite bond energy (34%). Consequently, the catalyst activity of propylene polymerization mainly depended on steric hindrance. The predictive ability of the model was successfully validated with a set of five electron donors which randomly selected from three different groups. Predictive R2 for the test set was 0.77, indicating good predictive ability of the model. The QSPR model provided the valuable information for the design of better electron donors for propylene polymerization. (C) 2014 Elsevier Ltd. All rights reserved.