Seven tight-lattice NEPTUN-III bottom-flooding experiments are analyzed by using the frozen version of RELAP5, RELAP5/MOD3.3/BETA. This work is part of the Paul Scherrer Institute (PSI) contribution to the High Performance Light Water Reactor (HPLWR) European Union project and aims at assessing the capabilities of the code to model the reflooding phenomena in a tight hexagonal lattice (which was one of the core geometries considered at the time for an HPLWR) following a hypothetical loss-of-coolant accident scenario. Even though the latest version of the code has as a default the new PSI reflood model developed by the author, which was tested and assessed against reflooding data obtained at standard light water reactor lattices, this work shows that for tight lattices, the code underpredicts the peak clad temperatures measured during a series of reflooding experiments performed at the NEPTUN-III tight-lattice heater rod bundle facility. The reasons for these differences are discussed, and the (possible) changes needed in the framework of RELAP5/MOD3.3 for improving the modeling of reflooding in tight lattices are investigated.