In the present investigation, a combination of docking and quantitative structure-property relationship (QSPR) approaches was applied to elucidate the host-guest interactions in β-cyclodextrin complexation with diverse set of organic compounds. Molecular docking was performed to find correct conformations of organic molecules in the cavity of β-cyclodextrin. The conformation with the lowest binding free energy was chosen to calculate molecular descriptors. Some additional descriptors relevant to characterizing the structural properties of inclusion complexes were also calculated and used in QSPR model building. Genetic function approximation technique was applied to choose the best subset of descriptors. The selected descriptors explain that the hydrophobicity, surface area and shape of guest molecules play important roles in the β-cyclodextrin complexation. The final QSPR model, based on multiple linear regression (MLR) method, was characterized by satisfactory statistical performance; calibration (R2c) and prediction (R2p) correlation coefficient of 0.83±0.02 and 0.78±0.07 respectively.