In a community survey, subjects and their informants were asked the same questions about memory and intellectual decline in the subjects. Subjects and informants both commonly reported cognitive decline, although in most cases the decline was not seen as interfering with daily life. However, when responses from subjects and informants were cross-tabulated, agreement was found to be poor. For subjects, reports of cognitive decline were correlated with anxiety and depression symptoms and with trait neuroticism. Subjects' reports were uncorrelated with age and only weakly correlated with cognitive test performance, indicating little validity. By contrast, informants' reports were correlated with the subjects' cognitive test performance and age, but also with the informants' own anxiety and depression symptoms. Although informants' reports have validity, they may also be contaminated by the informants' affective state.