We are pleased that our paper on the stress transfer within double-walled carbon nanotubes (DWNTs) in composites [1] is being read and promoting discussion in the field, but we do not agree with most of the comments by Lin and Wang upon the Communication.[2] We feel that they have missed some of the key concepts in the paper and, although they state that the data should be reinterpreted, they do not suggest an alternative approach.Lin and Wang [2] state that we have ''attempted to indicate that the capability of the outer wall of taking on the load transferred from the polymer matrix is reduced due to the poor wall-to-wall stress transfer''. However, we discuss clearly in our paper that they are two stress-transfer processes in the DWNTs, the polymer/outer-wall interface (k 0 ) and the wall/wall interface (k i ). We do not ever relate k 0 to the wall/wall interface, as Lin and Wang imply in their comment. This confusion between the two interfaces by Lin and Wang may help understanding why they are not satisfied with the paper. They do not have appeared to have recognized the extraction of k i as a parameter in the first experimental half of the paper, where k i is measured independently of the dispersion and polymer/nanotube interactions. k i is then used to predict the effective modulus of a nanotube, taking into account its full cross-sectional area, based upon the model of Zalamea et al. [3] Lin and Wang [2] also question the assumption that the Raman band shift rates depend linearly upon the Young's modulus of the nanotube. Lin and Wang argue that whilst it has been shown that this linear dependence exists for carbon fibres, it might not apply to single crystals such as nanotubes. This is a valid argument, but Ferrari et al. [4] have shown recently that this linear assumption also applies to graphene, which is the ultimate carbon single crystal.Throughout the paper we have assume that the atomic stiffness of a given wall is that of graphene. Thus, if everything was perfect in the composite, we would expect to see a similar band shift in each wall to that in graphene ($0.05 cm À1 GPa per %). However, we do not see such a shift. The reason for the low shift rate in the outer walls is due to the nonperfect polymer/ nanotube interface, which includes poor dispersion. We acknowledge in the paper that this poor dispersion is the reason why the DWNT outer walls have a lower shift rate than the walls of the single-walled carbon nanotubes (SWNTs). However, the fact we see a shift rate less than expected for a graphene-like lattice implies that the stress is being transferred poorly into the lattice; hence the effective modulus of this wall is reduced. We assume that the atomic stiffness is the same in both the inner and outer walls of the nanotubes. Therefore the difference in the Raman band shift of the inner and outer walls must be related to having less stress present in the inner wall than the outer. This difference in stress for a given strain allows us to measure directly the efficiency of the stress transfer ac...