Schachter (12) assumed that alcoholics do not handle their anxiety by-affiliation, and found some indirect support for this contention in his analysis of data provided earlier by Bakan (1). However, certain figures in the latter sample are incompatible, and the method of analysis was unrefined. Smart (15) supplied satisfactory data on family size and birth order among 242 alcoholics, and used appropriate techniques of analysis. He tested four hypotheses based on Schachter's work and assumption concerning alcoholics, and rejected three of them. If the corrected sizes of sibships in Smart's sample are compared with general population data on the sizes of completed families, this leads to rejection of the remaining hypothesis. The findings are compatible with the clinical observation that most male alcoholics are birds of a feather who tend to affiliate with other alcoholics. Smart (16) provided additional data on a large sample of female alcoholics, which showed a similar distribution of family sizes to his predominantly male sample, but also a marked excess of youngest-born members of large families. The latter finding has previously been reported in certain other large samples of psychiatric patients, but the evidence is conflicting and requires further elucidation.