Two hundred and ninety-one patients, with an average age of 69 in 1965, referred to a psychogeriatric clinic were studied. All but 20 were traced. 167 died and 104 remained alive. 73 were institutionalized. Medical, psychiatric and social data was available for all. Using mortality tables, we calculated the time of death for each patient. The group that exceeded their life expectancy was compared to the group that died prematurely. Significant positive correlations with longevity were self-referral, higher education, skilled work, independent income and absence of dementia. Females and orphans also lived longer. Living alone, dependency on children and conflict with one's spouse predisposed to institutionalization. Curiously, the hardships of being in wartime Europe and/or in a concentration camp increased life expectancy and mitigated against institutionalization. An attempt is made to correlate our findings with other studies and to explain our results.