SummaryForms of head‐scratching are described and classified. Indirect‐scratching differs from the direct chiefly in that it involves the positive movements of lowering one wing and bringing the foot up over it. Both methods of scratching exist on two main levels: basic‐scratching is a reflex‐like response to irritation, etc., on the bill or head, while extended‐scratching functions as part of the feather‐maintenance system. Indirect‐scratching has been recorded as an apparent displacement‐activity in some species.Possible evolutionary trends in head‐scratching are discussed. The two methods may have evolved independently from a more primitive method or successively one from the other. If the two methods evolved successively, then it is argued, contrary to the classical view of Heinroth, that indirect‐scratching evolved later than direct‐scratching, probably after, or concurrently with, the development of the extended function of scratching. Support is provided for this view by the distribution of indirect‐scratching in the bird‐kingdom and by the occasional use, by species in which indirect‐scratching is the rule, of a scratching method with some characters of the direct. The latter may be explained as either a reversion to the primitive method and/or the use of incomplete or wrongly co‐ordinated indirect movements. Variation in scratching within the avian family may be similarly explained and/or is due to incorrect taxonomy (e.g. in Parulidae and Timaliidae). As the method of scratching, generally speaking, is uniform in related birds up to at least the family level (in truly monophyletic groups), head‐scratching method is a potentially useful taxonomic character.The scratching movements of a number of species are described and discussed, including the peculiar scratching of penguins, and a number of new re