Some retrospective data collected in a survey carried out in 1967–68 show that the practice of breast-feeding declined markedly in Great Britain between the 1930s and the 1960s. Throughout, women higher up the social-educational scale breast-fed more than those lower down. Women marrying before the age of 20 (as well as those marrying at age 30 or older) tended to breast-feed less than those marrying in their twenties; for the former, the explanation was probably ‘social’. Needing, or choosing, to return to work soon after confinement was not, in general, an important inhibiting factor so far as breast-feeding was concerned. From the 1967–68 data, there seemed to be a negative association between the level of breast-feeding and birth order, but this may not have been ‘genuine’ since other data did not agree. There was a positive association between level of breast-feeding and family size, for family sizes up to four, but a drop for those with five or more children. This latter feature could be the result of a tendency for those with large families to arrive at this situation by way of short interbirth intervals, this in turn being associated with low levels of breast-feeding; the former might possibly reflect a positive link between women' capacity to breast-feed and their capacity to bear children.