Seven NEPTUN reflooding experiments with varying parameters flooding rate, single rod power, pressure and initial rod temperatures were simulated with the code RELAP5/MOD2, version 36.02, to assess the code, especially its reflood model. These calculations were performed with the specific objectives of evaluating the general prediction capability as well as specific problem areas of the RELAP5/Mod2 code in modelling boil-off and reflood behavior. First a study of the effect of the hydraulic and conduction nodalization to the results of the code was performed using a high and a low flooding rate experiment. After the choice of a proper nodalization, base calculations were done for all seven NEPTUN reflooding tests. The differences between code predictions and experiments are described and analysed. Implementing new correlations into the code and modifying or correcting existing correlations, for example for wall heat transfer or interphase friction, some of the weak points of the code during reflooding could be identified. These modifications were checked with all seven NEPTUN experiments. Additionally, two FLECHT-SEASET tests were simulated with the frozen version and also with the modifications mentioned above. (author) 60 figs., 4 tabs., 17 refs.