Experimental data from vented hydrogen deflagrations with concentrations from 15% to 21% are presented. The experiments were performed in a 27 m(3) cubic chamber with a vent area of either 0.78 or 1.76 m(2). Uniform and stratified mixtures were used in the test matrix, and the obstacle configuration also varied to achieve different volumetric blockage ratios. The purpose of this study was to investigate the effect of hydrogen concentration, vent area, non homogeneous mixtures and obstacles on the explosion venting for lean hydrogen-air mixtures, and to evaluate the performance of an analytical model for calculating the maximum reduced overpressure. The results show that the maximum overpressure measured inside the empty chamber increases from 3 kPa to 28.5 kPa as hydrogen concentration increases from 15.3% to 20.2%. The raised vent location may affect the pressure development during explosion venting, and a third pressure peak caused by the occurrence of the maximum flame surface area could appear in the pressure profile when a large vent is located on the upper part of the side wall. Higher volumetric blockage ratio leads to higher maximum overpressure and flame velocity, and the increment of each pressure peak with the blockage ratio is more pronounced for the back ignition. Furthermore, when comparing with the volume of unburnt gases, the obstacles have a greater effect on the flame. However, the obstructions have a limited effect on the non-homogenous hydrogen deflagration, whose combustion behaviour is governed by the maximum hydrogen concentration in the chamber. Molkov' best-fit model over-predicts the maximum reduced overpressure measured inside the empty chamber, but the predictions are relatively acceptable for the center ignition. (C) 2019 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.