The TaBoo SeArch (TBSA) algorithm [ Harada et al. J. Comput. Chem. 2015 , 36 , 763 - 772 and Harada et al. Chem. Phys. Lett. 2015 , 630 , 68 - 75 ] was recently proposed as an enhanced conformational sampling method for reproducing biologically relevant rare events of a given protein. In TBSA, an inverse histogram of the original distribution, mapped onto a set of reaction coordinates, is constructed from trajectories obtained by multiple short-time molecular dynamics (MD) simulations. Rarely occurring states of a given protein are statistically selected as new initial states based on the inverse histogram, and resampling is performed by restarting the MD simulations from the new initial states to promote the conformational transition. In this process, the definition of the inverse histogram, which characterizes the rarely occurring states, is crucial for the efficiency of TBSA. In this study, we propose a simple modification of the inverse histogram to further accelerate the convergence of TBSA. As demonstrations of the modified TBSA, we applied it to (a) hydrogen bonding rearrangements of Met-enkephalin, (b) large-amplitude domain motions of Glutamine-Binding Protein, and (c) folding processes of the B domain of Staphylococcus aureus Protein A. All demonstrations numerically proved that the modified TBSA reproduced these biologically relevant rare events with nanosecond-order simulation times, although a set of microsecond-order, canonical MD simulations failed to reproduce the rare events, indicating the high efficiency of the modified TBSA.