Among the flourishing Muroidea, the Eucricetodontinae constitute the most diversifi ed subfamily of the Eurasian Palaeogene. Early representatives, such as Atavocricetodon and Eucricetodon, are notably characterized by their simple and primitive dental morphology. As a result, it appears very diffi cult to discriminate these genera according to their dental morphology, and that is emphasized by their imprecise diagnosis. Here, we aim at challenging the taxonomic validity of Atavocricetodon, frequently considered as a grade of Eucricetodon. We chose to investigate the outline of fi rst upper molars for four species of Atavocricetodon and seven of Eucricetodon, to avoid confusions made by their occlusal dental characters. The outline analysis involves samples of ten molars, and was realized by using Elliptic Fourier Transforms. Incisor enamel microstructures were investigated as well in four species, because this dental character appears highly useful in rodents to solve taxonomic issues. Results revealed that if a signifi cant difference exists between the global dental shape of Atavocricetodon and Eucricetodon, outline morphospaces of the considered species generally overlap as displayed by the multivariate analysis and the hierarchical clustering. This is partly explained by the high intrageneric shape variation, which might correspond to different Oligocene lineages of Eucricetodontinae. Such variation is also due to Atavocricetodon paaliense, which could rather belong to another genus, the pseudocricetontine Kerosinia. More generally, there is no discrete character or real apomorphy which permits to distinguish these genera, because incisor enamel microstructure patterns are highly variable and a few types are shared by both genera. Atavocricetodon cannot thus be retained neither as a genus, nor as a morphosubgenus of Eucricetodon. Further studies, including much more fossil material, are needed to accurately identify the various lineages of Eucricetodon and to potentially defi ne new genera on a fair basis.