This paper attempts to examine the effect of the scaling of ground motion records on the seismic responses of base-isolated buildings by using the three scaling methods including the ASCE 7-10, acceleration spectrum intensity, SIa, and the spectral acceleration corresponding to the effective period of the base-isolated structure at the design displacement, S-a(T-D). The buildings had steel moment-resisting and X-braced systems and were isolated with lead rubber bearings (LRBs) at the base. Near-fault ground motions with Forward Directivity (FD), Fling-Step (FS) and No Pulse (No P.) characteristics as well as far-fault ground motion records were studied in this paper. The discrepancy in the seismic responses as well as the effect of the pulse period in the case of pulse-like ground motions was scrutinized for the three scaling methods. Also, the response spectrum analysis (RSA) and static method with the equivalent lateral force distribution prescribed in ASCE 7-10 were carried out. The results indicate that the RSA as well as the static analysis, prescribed in ASCE 7-10, underestimates the story shears of the base-isolated buildings. The underestimation increases in the case of the FF and No P. ground motions having a large amount of spectral acceleration in the higher modes region. Also, the underestimation increases with the increase in the height of the base-isolated buildings as well as in the case of the moment-resisting frames compared to the X-braced ones. Finally, by means of the results derived from the time history analyses, a formula was proposed to determine the lateral force distribution along the height of the base-isolated buildings with LRBs considering the characteristics of the isolator and the structure for different types of the ground motions.