A new modified version of the RELAP5/MOD2 computer code for the analysis of the reflood phase after a hypothetical large-break loss-of-coolant accident is developed. Various rewetting correlations are examined and compared with full-length emergency core heat transfer separate-effects and system-effects test (FLECHT-SEASET) experimental reflood data. The RELAP5 prediction of vapor temperatures is low in comparison with the data. The use of a new interfacial heat transfer between droplets and steam results in a reasonable prediction of vapor superheats. A revised dispersed flow film boiling correlation, which accounts for the enhancement of steam convective cooling by droplet-induced turbulence, is incorporated in the code. Comparison of the current results with data shows significant improvement in the prediction of clad temperature time histories over previous RELAP5 calculations.