Two main types of data are available to resolve phylogenies using fossils data: (1) stratigraphic ordering of taxa, and (2) morphological characters. In most phylogenetic studies dealing with ammonites, authors have given priority to the stratigraphic distribution of taxa. This practice is classically justified by the fact that the ammonite fossil record is frequently outstandingly good. In practice, the level of integration of stratigraphic and morphologic information in a single analysis depends on the confidence that authors have in the quality of data. Besides, many evolutionary concepts, which could differ over time and between authors (e.g. anagenesis, cladogenesis, iterative evolution), are added to these data to help infer phylogenetic relationships. As a result, phylogenetic hypotheses are based on eclectic methods which depend on the relative weight given to stratigraphic and morphologic information as well as on evolutionary concepts used. The validity of relationships proposed by previous authors is not dealt with in this paper. Instead, our goal is to draw attention to problems that these eclectic methods may cause, that is to say: (1) ammonites systematics is poorly formalised and (2) phylogenetic hypotheses as they are classically constructed are not rigorously testable. During the last 10 years, cladistic analysis has been applied to ammonites but is still unpopular among ammonitologists. However, studies have consistently shown that cladistics is not as unsuited a tool for ammonites phylogenetic reconstruction as is widely believed. Moreover, classical works open new questions about ammonite phylogeny and in particular, help to reappraise our view on the definition of morphological characters and their phylogenetic significance.