In a recent letter, 1 Si (313) rather than was used as a reference to determine the alloy composition and the in-plane strain built in the AlGaN buffer layers of a GaNon-Si heterostructure via reciprocal space mapping (RSM) of the nitride (10-14) atomic planes. 1 The authors claimed that this method is more accurate than currently favored method which uses GaN as the reference. Unfortunately, the crystallographic tilt, i.e., the angle intrinsically formed between III-nitrides (0002) and Si (111), is unintentionally ignored by the authors. Another flaw is that the authors assume the GaN, as the reference in currently favored method, to be completely relaxed. As a result, the accuracy comparison made by the authors is misleading.Crystallographic tilt has been observed in various GaNon-Si heterostructures grown by metalorganic chemical vapor deposition (MOCVD) on either unintentionally or intentionally miscut Si (111) substrates. [2][3][4] The unintentionally miscut, particularly those of large-diameter Si wafers, is up to 0.5 . We found that such an unintentionally miscut could lead to a tilt of $0.2 between the GaN (0002) and Si (111) atomic planes, contributed by the Nagai tilt, misfit dislocations, and/ or dislocation bendings. 2 However, the crystal orientation of the III-nitride layers always follows that of the first nitride buffer immediately on the Si substrate. Obviously, the tilt between nitrides (0002) and Si (111) has different components towards the three 120 -rotated Si (313) planes. As a result, the nitrides (10-14) RSMs carried out in different azimuths (120 -rotated) using Si (313) as references would lead to different alloy compositions and/or strains. The worst case is that the tilt is right in one of the three 120 -rotated azimuths and the (10-14) RSM is happened to be carried out in this orientation. In this case, a tilt of 0.2 can cause a change of 0.029 Å to the lattice constant a of relaxed GaN, which leads to an inaccuracy of De a ¼ 0.92% in its in-plane strain measurement. This inaccuracy is even larger than the largest in-plane strain (typically that of AlN) of the nitride layers in GaN-on-Si. 1,5 To experimentally demonstrate the inaccuracy when using Si (313) as the reference, we have measured nitride (10-14) RSMs from an AlGaN/GaN high electron mobility transistor (HEMT) structure grown by MOCVD on an onaxis 200-mm diameter Si (111) wafer (miscut of 0.5 ). The nitride stack has the similar buffer structure as that measured in Ref. 1. The HRXRD system (Panalytical X'PERT PRO MRD XL) and its setup employed in this demonstration are also the same as those used in Ref. 1. We carried out the nitrides (10-14) RSM measurements in the three 120 -rotated azimuths using Si (313) atomic planes of the Si (111) substrate as the references. Figure 1(a) presents the (10-14) RSM results, and the reciprocal points in each set of the RSMs were collectively shifted, following Ref. 1, by locating Si (313) to (Q X , Q Z ) ¼ (3.0067, 7.4412) nm À1 . The Al compositions [Fig. 1(b)] and the in-plane strains ...