A series of recent papers (Creak, 1961, 1963; Rutter, 1966; Brown, 1963; Lotter, 1967; Schain and Yannet, 1960) have provided evidence of degrees of cerebral dysfunction in infantile autism (Kanner, 1943) and other infantile psychoses. They have demonstrated that groups of cases of infantile psychoses satisfying broadly similar diagnostic criteria have in their backgrounds a variable frequency of cerebral insult and abnormal discharge in the EEC The diagnostic criteria in such series are of crucial importance. Imprecision and vagueness hamper comparisons between series. In childhood psychoses some investigators have excluded those cases in which there was any history or clinical evidence of organic features in an attempt to obtain a ‘pure’ group. This technique, while valid in delineating a syndrome, is handicapping to subsequent etiological study. For these reasons the present authors have used only age of onset and behavioural features as their ascertainment criteria it was a behavioural rather than an aetiological diagnosis.