The very elderly (over 80 years) have a relatively high level of disability and dependency, and this has been reported to be a greater problem in women. It has, therefore, been proposed that there may be little to gain from prolonging life of the very elderly, especially very elderly women. However, there is very little evidence to suggest that increasing survival leads to a higher prevalence of disability at a given age. Preventive measures that prolong life appear to postpone both morbidity and mortality. Moreover, there is little reliable evidence that an elderly woman is more dependent than a man of the same age. It is also suggested that there is a 'cap' on survival and, therefore, we should concentrate on preventing morbidity and compress it into the last few years of life. We consider whether or not the elderly are approaching a 'cap' on survival and think that they are not. We conclude that preventive medicine in the elderly should aim to prolong survival and prevent morbidity. It is uncertain whether morbidity can be prevented more than mortality or vice versa. The concept of compressing morbidity at the end of life is an attractive concept, but data are lacking to support this idea.