A recent theoretical treatment of intermolecular nuclear-spin relaxation that incorporates the effect of molecular rotation [K. Lendi, Phys. Rev. A 45, 7906 (1992)] is shown to be based on what appears to be a mistaken assignment of Euler angles. In particular, the conditional probability for reorientation of the intermolecular vector due to molecular rotation seems to be given incorrectly. Unsolvable integrals that are examined by Lendi are derived without considering the reflecting wall boundary conditions that are requisite for the problem of translational diffusion.