This paper describes shaking table tests of three eight-story building models: all are masonry structures in the upper stories, with or without frame-shear walls of one- or two- stories at the bottom. The test results of damage characteristics and seismic responses are provided and compared. Then, nonlinear response analyses are conducted to examine the reliability of the dynamic analysis. Finally, many nonlinear response analyses are performed and it is concluded that for relatively hard sites under a certain lateral stiffness ratio (I.e., the ratio of the stiffness of the lowest upper masonry story to that of the frame-shear wall story), the masonry structure with one-story frame-shear wall at the bottom performs better than a structure built entirely of masonry, and a masonry structure with frame-shear wall of two stories performs better than with one-story frame-shear wall. In relatively soft soil conditions, all three structures have similar performane. In addition, some suggestions that could be helpful for design ofmasomy structures with ground story of frame-shear wall structure in seismic intensity region VII, such as the appropriate lateral stiffness ratio, shear force increase factor of the frame-shear wall story, and permissible maximum height of the building, are proposed.