Abstract
177 people aged 65 or over, chosen at random from larger representative samples of elderly people living at home in Newcastle upon Tyne, were given the Weschler Adult Intelligence Scale (WAIS) or a shortened form of it, and followed up for 7 years or till death. Discriminant function analysis showed that the power of the test score to predict death occurring within 2 years was not explained away by its correlations with age, sex, social class or physical disability. Exclusion of clinically diagnosed chronic brain syndromes reduced but did not abolish the relationship found to exist between test score and outcome. The ascertainment of impaired cognitive functioning has important applications in the assessment of prognosis and in the planning of care of elderly people.