Does the application of seismic Born theory, as opposed to simpler ray theory, lead to an improvement in tomographic images of the Earth? In recent publications, Montelli et al. (2004a, 2004b) and van der Hilst and de Hoop (2005) among others have expressed opposite opinions. We propose a quantitative approach to the comparison of tomographic images, which we apply to the case of surface-wave phase velocity maps derived with Born vs. ray theory.