The taxonomic treatment of trace fossils needs a uniform approach, independent of the ethologic groups concerned. To this aim, trace fossils are rigorously defined with regard to biological taxa and physical sedimentary structures. Potential ichnotaxobases are evaluated, with morphology resulting as the most important criterion. For trace fossils related to bioerosion and herbivory, substrate plays a key role, as well as composition for coprolites. Size, producer, age, facies and preservation are rejected as ichnotaxobases. Separate names for undertracks and other poorly preserved material should gradually be replaced by ichnotaxa based on well-preserved specimens. Recent traces may be identified using established trace fossil taxa but new names can only be based on fossil material, even if the distinction between recent and fossil may frequently remain arbitrary. It is stressed that ichnotaxa must not be incorporated into biological taxa in systematics. Composite trace fossil structures (complex structures made by the combined activity of two or more species) have no ichnotaxonomic standing but compound traces (complex structures made by one individual tracemaker) may be named separately under certain provisions. The following emendations are proposed to the International Code of Zoological Nomenclature: The term ‘work of an animal’ should be deleted from the code, and ichnotaxa should be based solely on trace fossils as defined herein.