In this study with 161 elderly patients (mean age: 80 years), recently admitted to a psychogeriatric nursinghome, one-year survival was correlated with ratings derived from the Beoordelingsschaal voor Oudere Patiënten (BOP), i.e. Rating Scale for Elderly Patients. This rating scale, which has been adapted from the Stockton Geriatric Rating Scale and put into use in the Netherlands since 1971, has proven to be a reliable and valid method to study the behaviour of psychogeriatric patients. Variables associated with one-year survival (62 patients died within one year) were: helplessness, physical disability and inactivity. After partial correlation analysis controlling for the effects of sex and age, the following items showed a significant relationship to survival: eats without help, is not incontinent, does not need special support to prevent falling out of chair or bed, walks without aid, is cooperative, enters into conversation, is not incontinent and/or restless at night. In this study 78 patients could be tested by a newly devised method which consists of presenting a series of slides to assess spontaneous speech, naming, reading, comprehension, free recall, delayed recognition, visual perception and clockreading. These cognitive functions however did not show any, significant, relationship with survival, though the 25 patients dying within one year were significantly more disabled and more depressed than the 53 patients who survived more than one year. Ratings of psychomotor function thus proved to be of more prognostic value than assessments of cognitive functions.