Dedicated subchannel analysis codes have been employed to subchannel analyses. However, the recent development of thermal-hydraulic system codes with advanced two-phase flow models and three-dimensional components has resulted in more accurate and precise prediction of multiphase phenomena in nuclear reactor systems. The accurate prediction of the void fraction is one of the most important factors in subchannel analyses. This paper aims at assessing the applicability of three different system codes to the prediction of the void fraction against NUPEC experiment used for OECD/NRC PSBT benchmark. In this study, all cases for single channels and bundles in the benchmark database have been analyzed by using TRACE V5.0 Patch 5, MARS-KS 1.4, and RELAP5/MOD3.3 Patch 5, respectively. It is concluded that the applied system codes have a systematic difficulty in void fraction prediction under low void fraction conditions and additional assessment should be performed for low void fraction conditions including subcooled boiling regime.