A sample with a composition of 95 wt% Mg-5 wt% NbF5 (named Mg-5NbF(5)) was prepared by reactive mechanical grinding using Mg instead of MgH2 as a starting material. Its hydriding and dehydriding rates were then measured under nearly constant hydrogen pressures. The activation of Mg-5NbF(5) was not required, and Mg-5NbF(5) had an effective hydrogen storage capacity, which was defined as the quantity of hydrogen absorbed for 60 min, of 5.50 wt%. At the first cycle (n = 1) at 593 K, the sample absorbed 4.37 wt% H for 5 min and 5.50 wt% H for 30 min under 12 bar H-2, and desorbed 1.03 wt% H for 5 min, 4.66 wt% H for 30 min, and 5.43 wt% H for 60 min under 1.0 bar H-2. Reactive mechanical grinding of Mg with NbF5, which formed MgH2, MgF2, NbH2, and NbF3 by the reaction of 11 Mg + 7NbF(5) + 3H(2) -> MgH2 + 10MgF(2) + 2NbH(2) + 5NbF(3), is considered to create defects, to produce reactive clean surfaces, and to reduce the particle size of Mg. The XRD pattern of Mg-5NbF(5) dehydrided at n = 3 revealed Mg, small amounts of beta-MgH2 and MgO, and very small amounts of MgF2 and NbH2. An increase in the dehydriding rate of Mg-5NbF(5) was attempted by adding Ni to Mg-5NbF(5). Mg-5NbF(5) had higher initial hydriding and dehydriding (after the incubation period) rates and a larger effective hydrogen storage capacity than Mg-10NbF(5), Mg-10MnO, and Mg-10Fe(2)O(3), which were reported to have quite high hydriding rate and/or dehydriding rate. Copyright (C) 2014, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.