QSAR investigations of lipophilicity (XLOGP3) and biological activity (IC50) values of some Doxazolidine derivatives were conducted using combinations of multiple linear regression (MLR) and artificial neural network (ANN) modeling methods and three different optimization techniques including simulated annealing (SA), genetic algorithm (GA) and Imperialist Competitive algorithm (ICA). In addition CORAL software was used to correlate the lipophilicity and biological activity to the structural parameters of the drugs. The obtained results were compared and GA-ANN and ICA-MLR combinations showed the best performance with regard to the correlation coefficient (R2) and root-mean-square error (RMSE). The most effective descriptors extracted from lipophilicity and biological activity studies were presented and discussed. From GA-ANN method, the most important physico-chemical descriptors were found to be minimum value in atomic Sanderson electronegativities and maximum value in Squared Moriguchi Octanol-Water partition coeff.(log P ˆ2) descriptors.ICA-MLR method suggests the maximum value in polarizibility, electrotopological state and atom van der Walls volume as the most important physicochemical descriptors.It was concluded that QSAR study and Monte Carlo method can lead to a more comprehensive understanding of the relation between physico-chemical, structural or theoretical molecular descriptors of drugs to their biological activities and Lipophilicity. Manuscript profile