The present paper reports the results of an experimental study of detonation limits for H-2/O-2/Ar mixtures. Two stoichiometric mixtures (2H(2) + O-2 + 3Ar and 2H(2) + O-2) in three different diameter round tubes (D = 1.8, 4.6 and 10.9 mm) were tested. The choice of the mixture represents those considered as "stable" with a regular cellular pattern and "unstable" with an irregular cellular pattern. Detonation velocity was measured by ionization probes spaced at regular intervals along the small tubes. Consistent with previous findings, the present results show that well within the limits the detonation wave in hydrogen mixtures propagates at a steady velocity close to the theoretical Chapman-Jouguet (CJ) value. With decreasing initial pressure, the velocity deficit increases. It is found that the detonation velocity decreases with decreasing tube diameter, which is a result of the wall boundary layer effect being more prominent for smaller tubes. At the limiting pressure, the steady velocity deficit for both tested mixtures in three different diameter tubes is about 15-18%. Velocity deficits were also estimated theoretically using the Fay model. For the mixture of 2H(2) + O-2+ 3Ar, good agreement is found between the theoretical prediction and the experimental result of detonation velocity. For the mixture of 2H2 + 02, however, the theoretical prediction deviates from the experimental measurement. The latter thus suggests that, apart from losses due to the flow divergence caused by the boundary layer effect, instabilities are also significant for the detonation propagation and failure in 2H(2) + O-2 mixture. Lastly, at the limits, the value of D/A is found to agree well with that for the onset of single-headed spin detonation. Thus, it can be concluded that the single-headed spin criterion can also be used for defining the detonation limits for hydrogen mixtures. (C) Copyright 2016, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.