In a recent paper in this journal, Rudyak, Dubtsov, and Baklanov (2009) presented results of measurements of the penetration of nanoparticles with diameters from 3.5 to 84 nm at temperatures from similar to 300 to 600 K through a set of wire screens, from which they inferred diffusion coefficients. They argued that the formulation typically used for C. the Cunningham correction that accounts for non-continuum effects on the diffusion of nanoparticles, is not valid for temperatures greater than similar to 300 K, and they proposed a modification of this formulation which depends on both temperature and particle size. It is shown here that this modification produces unphysical results in that it yields negative values of the momentum accommodation coefficient. A likely reason for their results is that they used a polydisperse size distribution, for which the main contribution to the measured penetration would be from particles at sizes far from those attributed to them. (C) 2010 Elsevier Ltd. All rights reserved.