Investigations on the 1999 Taiwan Chi-Chi earthquake indicate that most of the bridge columns experienced none to-minor damages except for those bridges directly crossed by the ruptured faults. Studies show that the functions of the bearing system, including non-bolted rubber bearings, shear keys and restrainers, plays an important role for the performance of the bridges during the earthquake. One of the reasons for this unexpected performance is the way of bridge bearing construction in Taiwan. This paper aims at understanding the effect of bridge bearing systems and providing a simplified method to estimate the maximum displacement demand. Experimental studies on friction coefficients of rubber bearing on non-shrink cement mortar or concrete surface were first carried out. A SDOF bridge system with rubber bearings were also studied under pseudodynamic tests and verified by a numerical model. Test results show that, in a quasi-static state, the friction coefficient of rubber bearing is about 0.2 on three different contact surfaces. The numerical model is adequate to obtain the global displacement demand when the rubber bearing is simulated by a frictionpendulum element with 15-20% reduced shear stiffness during large shear deformation. Based on the experimental results, this study proposes a simplified ATC-40 method considering the equivalent friction damping to determine performance point. The numerical result of a SDOF bridge predicts the maximum displacement of the bridge deck well when compared with the mean value from 14 nonlinear time-history dynamic analyses. Results of this study help to understand the damage that may occur to the rubber bearings and bridges columns under major earthquake ground motions.