ET(A) subtype selective antagonists constitute a novel and potentially important class of agents for the treatment of pulmonary hypertension, heart failure, and other pathological conditions. In this paper, 60 benzodiazepine derivatives displaying potent activities against ET(A) and ET(B) subtypes of endothelin receptor were selected to establish the 3D-QSAR models using CoMFA and CoMSIA approaches. These models show excellent internal predictability and consistency, external validation using test-set 19 compounds yields a good predictive power for antagonistic potency. Statistical parameters of models were obtained with CoMFA-ET(A) (q2 = 0.787, r2 = 0.935, r2pred) = 0.901), CoMFA-ET(B) (q2 = 0.842, r2 = 0.984, r2pred = 0.941), CoMSIA-ET(A) (q2 = 0.762, r2 = 0.971, r2pred = 0.958) and CoMSIA-ET(B) (q2 = 0.771, r2 = 0.974, r2pred = 0.953) respectively. Field contour maps (CoMFA and CoMSIA) corresponding to the ET(A) and ET(B) subtypes reflects the characteristic similarities and differences between these types. The results of this paper provide valuable information to facilitate structural modifications of the title compounds to increase the inhibitory potency and subtype selectivity of endothelin receptor.