In catalyst activity testing, micro-packed-bed reactors are the most commonly used devices. However, the small particle sizes (typically 0.05-0.2 mm) inherent with these microreactors in multiphase systems do exhibit special hydrodynamics characteristics, that are not generally recognized. Cold three-phase model investigations show that in micro-packed-bed reactors segregated flows of gas and liquid occur in a broad range of conditions. This situation is dramatically different from the hydrodynamics in industrial-scale trickle-bed reactors. Since in a broad range of conditions the gas flow follows preferential pathways through the micro-packed bed, the average thickness of the liquid layers is relatively large. Because the transport through these layers occurs mainly by molecular diffusion, the observed reaction rate potentially will be limited, resulting in poor radial dispersion. Even for a rather slow reaction as the hydrodesulfurization of dibenzothiophene (HDS), radial mass-transfer limitations were observed. This counterintuitive result is due to the combination of (i) the presence of apparent stagnant liquid layers with several particle diameters thickness, (ii) the low rate of radial transport, and (iii) the strong inhibition by the reaction product H2S. A general criterion, analogous to the Weisz-Prater criterion for internal diffusion limitations, is proposed for the estimation of the influence of poor radial dispersion in catalyst performance testing in micro-packed-bed reactors. For use in practice, it is advised to start with a worst-case scenario and dependent on the outcome to relax this strict criterion. (C) 2015 Elsevier B.V. All rights reserved.