Large differences have been found between the Isle of Wight (IOW) and a (former) inner London borough (ILB) with respect to the rates of emotional (neurotic) disorders, conduct disorders and specific reading retardation in ten-year-old boys and girls (Rutter et al., 1975; Berger et al., 1975). This paper presents preliminary considerations on possible reasons for these differences. Any explanation must account for the fact that the differences in the rates apply to three rather disparate disorders. The epidemiological characteristics and psychosocial correlates of emotional disorders in childhood differ quite markedly from those of conduct disorders, and it appears highly likely that the causes of the two disorders are different (Rutter, 1965, 1970a; Rutter, Tizard and Whitmore, 1970). Although specific reading retardation overlaps markedly with conduct disorders, to a considerable extent it seems to have a separate set of aetiological influences (Rutter and Yule, 1973). Furthermore, there is little association between specific reading retardation and emotional disorder in childhood. It is also notable that the higher rates of disorder in London compared with the Isle of Wight apply just as much to girls as to boys, in spite of the marked differences between the two sexes in the correlates of psychiatric disorder (Rutter, 1970b).