The occurrence of a phylogenetic signal in bone microstructure and histology, and their use in taxonomy, has been debated. We give here a simple example that shows how the addition of histological data can resolve an issue of taxonomy. Specimen MNHN.F.PPI 110 is an isolated rib found in the Late Miocene/Early Pliocene of the Pisco Formation, Peru. It was assigned to an unknown and unnamed dugongid, based on its gross morphology and its microstructure. The description of its histological features suggests that the first attribution was erroneous. The integration of the three levels of organization reveals that it can rather be assigned to the aquatic sloth Thalassocnus , either T. carolomartini or T. yaucensis . This result documents the case of an extreme degree of osteosclerosis, unprecedented for a non-sirenian taxon. Additionally, it implies a revision of the spatial and temporal distribution of sirenians.