This study develops a statistical relationship between the peak displacement of seismic isolation systems obtained from nonlinear time-history dynamic (NTH) analysis and the peak displacement predicted from an equivalent linear force procedure (ELF). Firstly, the effect of isolators’ mechanical property and excitation type and amplitude on the ratio between the peak displacement of an isolation system calculated from ELF procedure and the peak displacement determined from NTH are evaluated. The ELF procedure in ASCE/SEI 7–16 is employed to estimate the ELF peak displacement. The input ground motions for bidirectional NTH analysis are scaled following this code. The investigation of 540 numerical models subjected to four levels of excitation amplitude, represented by 50 pairs of ground motion, including near-field and far-field motions, reveals that both mechanical property of an isolation system and excitation type and amplitude affect the displacement ratio, where the post-yield stiffness of the isolation system and ground motion type have stronger influence than other parameters. These influences are then statistically processed. Based on the investigated data, a novel practical equation accounting for these parameters is proposed to predict the displacement ratio at different reliability levels.