Hyper-Kamiokande (HK) is a next generation multi-purpose neutrino and nucleon-decay experiment scheduled to begin operation in 2027. The experimental setup of HK will consist of the neutrino beam line and a suite of near detectors at J-PARC, and a far detector of the same name, HK. The far detector HK, filled with 260 kton of pure water is able to detect a Cherenkov light from the relativistic charged particles produced in the (anti)neutrino interactions. To distinguish neutrino and antineutrino interactions in HK, a neutron signal can be used. This is especially important in the Diffused Supernovae Neutrinos Background search, which is one of the physics targets of HK. Using neutron signals allows HK to highly suppress background events and to improve the signal sensitivity. Furthermore, the neutron tagging largely reduces neutrino-induced backgrounds with neutrons in the nucleon decay search where nucleon decays rarely accompany neutrons. In the pure water detector, the generated neutron is thermalized in water and captured by a hydrogen nucleus. The hydrogen nucleus emits a 2.2 MeV gamma-ray that is seen as a delayed signal of the prompt Cherenkov events eventually. In this study, we simulated the neutron signal and one of the dominant backgrounds, created by Rn, which might mimic the neutron signals, and evaluated tagging efficiency in HK. The neutron tagging algorithm and the results will be presented.