Human umbilical vein endothelial cells are involved with endothelial cell function in human beings. Aurones exhibit good anticancer activities against human umbilical vein endothelial cell lines and hence provide for a demanding and potential area of development of aurone derivatives as anticancer agents. In the present research an attempt has been made to correlate the anticancer activity of the aurones to the physicochemical descriptors through QSAR studies. kNN- MFA and other advanced methods of analysis were opted for a set of seventeen aurone derivatives to figure out the correlationship between steric and electrostatic descriptors. In 2D-QSAR, amongst the several models obtained, statistically significant model 1 was selected with r2=0.9095, q2=0.4682 and pred_r2=0.6132 with four descriptors Ssss-N count, SdSc-count, T_F_2_1 and StSc-count, which showed significant contributions in the model. The model was found to be significant within the corresponding limits of internal and external validation parameters. In 3D-QSAR, statistically significant model 2 was obtained using simulated annealing algorithm, with the corresponding q2 and pred_r2 values 0.7101 and 0.6304 respectively, showing internal and external predictivities 71% and 63% respectively. The four descriptors which contributed significantly to the model were E_299, S_851, S_564 and E_760. The models were found to show good statistical and predictive significance which can further be used for guiding ligand modification for the development of potential anticancer agents.