In earlier distance geometry related three-dimensional quantitative structure-activity relationships (Ghose, A. K.; Crippen, G. M. J. Med. Chem. 1984, 27, 901) the interactions of the ligand atom or group with the receptor site were evaluated empirically by using mathematical optimization techniques, without considering their physicochemical properties. In the present work we show how to use various physicochemical parameters in our three-dimensional receptor mapping. We have developed a model for E. coli DHFR using the inhibition data of 25 pyrimidines and 14 triazines. It gave a correlation coefficient of 0.893 and standard deviation of 0.530. It successfully predicted the binding data of five pyrimidines and five triazines.