T HE recent paper by Nan et al. 1 was read with considerable interest. Further contributions to advance the understanding of the complex subject of the thermal conductivity of composites are always welcome.Nevertheless, I wish to take this opportunity to express my concern in regard to the description of our earlier work on the same topic in less-than-positive terms not usually encountered in the open literature. For unknown reasons, we were not given the opportunity to provide our response as part of the usual peerreview process.For instance, in referring to the study of Russell et al. 2 of the thermal conductivity of silicon carbide whisker-reinforced silicon nitride (SiC(w)/Si 3 N 4 ), Nan et al. 1 stated that the effect of an interfacial thermal barrier was "totally neglected."In this regard, the experimental data of Russell et al. 2 do not provide any direct or indirect evidence for the existence of an interfacial thermal barrier. Furthermore, at the time their work was conducted, it was not generally recognized that a thermal barrier would exist at interfaces in intimate contact at ambient temperature. It was generally thought that interfacial thermal barriers were of interest only to the low-temperature physicist. 3,4 In this respect, the results of the study of Garrett and Rosenberg 5 are quite relevant. In this study, the effect of an interfacial thermal barrier on the thermal conductivity of epoxy-resin/powder compacts could be detected only at temperatures Ͻ10 K. Certainly, these findings would lead anyone to conclude that the basic mechanism(s) responsible for the existence of an interfacial thermal barrier would be absent at ambient (i.e., room) temperature. Furthermore, the extensive literature on the effect of surface roughness on heat transfer between engineering components (see, for instance, the review papers by Cooper et al. 6 and Lambert and Fletcher 7 ) generally focuses on the heat transfer over those areas where direct contact is absent (perfect thermal contact is assumed to occur over those area in direct contact).Nan et al. 1 also stated that the values obtained by Russell et al. 2 for the SiC(w) thermal conductivity represent "rough estimates." In fact, these values were "calculated" based on the assumption of the applicability of the theory of Raleigh. 8 As we now know, if an interfacial thermal barrier were present and the whiskers were oriented perpendicular to the heat flow, the values for the whisker thermal conductivity obtained in this manner represent lower bounds.A brief historical perspective of the understanding of the role of interfacial thermal barriers in composite thermal conductivity seems worthwhile. The first (as far as I am aware) theoretical analysis of this subject was presented by Benveniste and Miloh 9 late in 1986, immediately followed in 1987 by the study of Benveniste 10 and the independent studies of Chiew and Glandt 11 and Hasselman and Johnson. 12 The latter study was completed and submitted well before the publication of the theory of Benveniste and Miloh. 9...