In this paper, preliminary evaluation of RELAP5 code was made on its application to the design of second side passive residual heat removal system(PRHR) by comparing to the experimental data. Because the theoretical basis of RELAP5 code was one-dimensional flow hypothesis, RELAP5 code was inadequate in the simulation of natural convection heat transfer in accident cooling pool during the initial stage of PRHR. In later-period of PRHR, the heat transfer mode in accident cooling pool was nucleate boiling based, and pool boiling had absolute predominance in the heat transfer quality, and the calculation results fitted the experimental data quite well. RELAP5 code can be basically applied to the steady state performance analysis of second side passive residual heat removal system.