Compared with cereals, evolutionary studies of food legumes are of comparatively recent date. The publication in 1965 by Kaplan of an account of Phaseolus bean archaeology and domestication marks a watershed. Since this date, much has been published, and an attempt will be made to review progress and to define problems still remaining. The student of legume crop evolution is concerned initially with the location and timing of domestication and how the crop has subsequently changed under domestication. To this end, it is therefore necessary to collate phytogeographical, archaeological and morphological evidence in the first instance and supplement this with experimental taxonomic analysis of existing crops and their closest wild relatives. Linguistic evidence can also provide valuable information on the dissemination of crops and sometimes the chronology of domestication. The best evidence for the latter is usually the radiocarbon dating of archaeological plant remains; these, however, are not always available. Area of origin is usually best deduced from phytogeographical considerations; useful supporting evidence is sometimes available from archaeological material. Where there is an apparent conflict between the simplest interpretation of phytogeographic and archaeological evidence, most authorities would be inclined to prefer the phytogeographic interpretation. The major sources of information on domestication can therefore be considered in two broad classes-the biological and the archaeological (together with evidence from related disciplines such as anthropology and linguistics). The immediate objectives of crop evolutionary studies are in the first place to find, if possible, a wild counterpart of the cultivated plant, to determine the nature of the changes which have occurred under domestication and elucidate their genetic control where practicable. The search for close wild relatives of crop plants can be surrounded with uncertainties. The closest wild relative sought may be extinct; divergence may have been so considerable between cultigen and wild relative that a connection may be difficult to establish; the cultigen may be of hybrid origin (cf. wheat) and no one wild form can be considered as progenitor; a form close to the progenitor may still exist in the wild but have remained uncollected or unrecognized as a wild counterpart of a crop plant. There are two further situations which can arise: the wild and domesticated forms may in fact have diverged so little that there is no difficulty in establishing the connection, or divergence may have occurred but not to such an extent as to obscure the close relationship between wild counterpart and cultigen. It is the determination of the closeness of the relationships between cultigens and their wild counterparts which is perhaps the major preoccupation of those studying crop plant evolution at the present time. The three biological approaches which have been particularly useful are: