High mortality rates among schizophrenic patients from infectious diseases, particularly tuberculosis, pneumonia and gastro-enteritis, reported for periods up to the 1940s were shown not to be specific for schizophrenia, but were characteristic of the mental hospital population as a whole (Alstrom, 1942). Studies covering more recent times confirm the decline and virtual disappearance of mortality from tuberculosis (Baldwin, 1979), but an extensive literature continues to emphasise the relatively high mortality of the mentally ill, including those defined as schizophrenic (Innes & Millar, 1970; Tsuang & Woolson, 1977), and more recent record linkage studies (Herrman et al, 1983; Allebeck & Wistedt, 1986) have continued to show an excess of both natural and unnatural deaths. Long follow-up studies of reasonably large groups of well documented cases are relatively uncommon in this area and therefore the 532 cases in the Harrow study were carefully followed up from the point of view of the occurrence and cause of death.