Assessment of suicide risk is an important and complex issue. We know that a considerable proportion of those who kill themselves seek help before they do so (Robins et al, 1959; Barraclough et al, 1974), and it is therefore essential both to recognise indices of suicide risk and to react appropriately to them. This means being sensitive to many personal, social and environmental factors, rendered more problematic by the fact that each individual is uniquely responsive to life events. The traditional approach, based on clinical epidemiological considerations, has long identified such factors as depression, alcoholism, increasing age, social isolation, physical illness and significant loss as important correlates of suicide, but there is now increasing evidence of other important factors which are frequently associated with suicide. Patients who commit suicide exhibit a wide range of behavioural problems which may hamper recognition of underlying suicide risk (Seager & Flood, 1965); and young adults with a history of chronic social turmoil are highly represented amongst suicides (Ovenstone & Kreitman 1974).