Being a task of establishing spatial correspondences, medical image registration is often formalized as finding the optimal transformation that best aligns two images. Since the transformation is such an essential component of registration, most existing researches conventionally quantify the registration uncertainty, which is the confidence in the estimated spatial correspondences, by the transformation uncertainty. In this paper, we give concrete examples and reveal that using the transformation uncertainty to quantify the registration uncertainty is inappropriate and sometimes misleading. Based on this finding, we also raise attention to an important yet subtle aspect of probabilistic image registration, that is whether it is reasonable to determine the correspondence of a registered voxel solely by the mode of its transformation distribution.