Several surveys of mental illness in husband-wife pairs have all concluded that sick couples occur in the general population with a frequency significantly in excess of the number to be expected by chance (Penrose, 1944; Gregory, 1959; Kreitman, 1962; Ryle and Hamilton, 1962; Pond et al., 1963; Nielsen, 1964). The statistical methods employed in such investigations have been criticized by Hall et al. (1971), but these authors' arguments appear to suggest that the relative excess of conjoint illness has, if anything, been under-estimated.