Vaillant's pioneer work on Compositae systematics has been generally dismissed as unimportant for nomenclatural purposes, being pre-Linnaean. However, a German translation of it was published in 1754, about one year after the starting-point date of botanical nomenclature. In order to assess its possible impact on generic nomenclature, a thorough analysis of Vaillant's concepts, particularly at the level of genus and species, is presented here. Vaillant subdivided the Compositae into three "classes", essentially equivalent to Cardueae, Cichorieae and the remaining tribes, and further into "sections" (14), genera (107) and species (1144). By the study of literature and, to a minor extent, herbaria, 852 of the species (>74 %) have been identified, and the identifications are listed in an Appendix. The definition and nomenclatural status of all genera has been verified and their names typified, whereby Vaillant's views and their impact on early Compositae systematics are clarified. In practical terms, the positive and negative effects of the nomenclatural restoration of Vaillant's names are balanced. It is not advisable to blacklist the work as a whole, because its negative effects can be taken care of effectively by just six conservation/rejection proposals (made separately). Among the positive aspects are availability of a suitable name (Rhaponticoides) for a genus that had become nameless (due to conservation of Centaurea L. with a new type); resurrection of Rhaponticum to settle a long lasting and unresolved nomenclatural controversy; and redemption of the otherwise illegitimate but widely used name Volutaria. As an aside, four of the names considered to be correct for Vaillant's taxa, which did not so far exist in the appropriate combination, are validated (in the genera Cullumia, Hymenolepis, Symphyotrichum and Tanacetum).