Tooth attachment and implantation are two classical descriptors of dental anatomy. Tooth attachment distinguishes between teeth that are either fused to the jaw by bone, or suspended within a socket by a periodontal ligament. Tooth implantation describes the geometry of this attachment and has been broadly divided into acrodonty, pleurodonty, and thecodonty. Among extant amniotes, only mammals and crocodilians are considered truly thecodont, because they possess a complex attachment system that includes a periodontal ligament and true tooth sockets. These two amniote groups diverged from a common ancestor over 300 million years ago and are thought to have evolved thecodonty independently. This view has recently come under a great deal of scrutiny with the discovery of complex tooth attachment systems, including the presence of a ligamentous tooth attachment in numerous non-mammalian, non-crocodilian amniotes. This has spurred debate and inconsistencies over the conventional usage of tooth attachment and implantation categories and their evolutionary significance. We provide a comparative histological approach for describing tooth attachment and implantation in an exemplary, traditionally thecodont taxonomic group: the non-avian dinosaurs. The comparisons between theropod, hadrosaurid, and ceratopsid teeth show that all dinosaurs have ligamentous tooth attachments composed of identical dental tissues to those in mammals and crocodilians, but they show a diverse array of tooth attachment geometries, replacement modes, and bone architectures supporting the dentition. The methodology we follow allows researchers to tease apart phylogenetically and functionally significant features of tooth attachment and implantation that could be used in future studies.