Predictions by RELAP5/MOD2 and RETRAN-02 are compared on a model of the Westinghouse model E steam generator, which is a U-tube steam generator with an integral preheater section. The model is the result of a detailed nodalization study performed with RETRAN to determine the minimum number of nodes (or control volumes) required in the secondary side to model the response of steam generator water level and primary-side exit (cold-leg) temperature during startup testing and operational transients. Five transients are used as forcing functions to generate the response of the steam generator. These transients were selected based on providing both nominal and severe forcing functions on the heat removal capability of the secondary side. The steam generator transients investigated were (a) loss of feedwater, (b) turbine trip, (c) decrease in load demand, (d) increase in load demand, and (e) decrease in inlet feedwater temperature. Steam line exit mass flow rate, secondary-side liquid mass inventory and water level, and primary-side cold-leg temperature predictions are compared with the RETRAN-02 results. Reasonable comparisons are obtained between the RELAP5 and RETRAN code predictions, as well as qualitative behavior of simulation experiments.