Abstract: The startup experiment data and steady-state experiment data, which were required from the principle experiment platform of the passive residual heat removal system that is one tenth of its actural size, were used to evaluate RELAP5/MOD3.2 code. The results show that the effect of the steam velocity on the steam condensation heat transfer coefficient in RELAP5/MOD3.2 code for this experiment system is under-estimated. So the steam condensation heat transfer coefficient in the vertical pipe is a little-small and the calculated and experimental results are totally different. Therefore, the condensation heat transfer model in the vertical pipe was modified and the calculated results were consistent to experimental results. The research result indicates that it is necessary to modify the condensation heat transfer model when it comes to the analysis of the similar passive residual heat removal system.