A review of the literature has made it clear that there is still little insight in the prevalence of personality disorders in the elderly. Prevalence rates seem to decrease only in the group of cluster B. However, no longitudinal studies continuing in old age have been carried out. In this publication explanations are given for the low prevalence rates, such as the course of personality disorders, cohort effects and diagnosing. There seems to be little evidence that shortcomings in diagnosing are the main cause of the low rates. Problems in the diagnostic process are illustrated and some recommendations are made in order to adjust axis II of DSM to the lifecircumstances of the elderly. It would be appropriate to make it possible to diagnose 'personality disorder in remission'.