the present research was designed to investigate the relationships among; a) depression, b) the quantity of social interaction and, c) the quality of social interaction in the elderly. twenty-three residents aged 72 to 98 of a rural senior citizens home were tested. the subjects were administered the zung self-rating depression scale and flavell's (1975) measure of visual perspective-taking. the quantity of social interaction was measured by the subject's self-reports of their frequency of interaction with various groups of people and the number of persons they talked to most frequently. the measure of the quality of social interaction was derived from a round robin procedure based on the subjects' reported self-disclosure to peers. this enabled the identification of subjects who had relationships characterized by reciprocal intimate disclosures - or confidants. the analysis provided partial support for the validity of the quantity and the quality of social interaction measures. consistent with previous research, depression was negatively correlated with one measure of quantity of social interaction, the number of peers with whom the subject frequently talked. contrary to expectation, perspective-taking was not significantly correlated with the measures of quantity of social interaction or the measures of reciprocal disclosure. contrary to the quality of social interaction research, the intimacy of the subject's disclosures and the intimacy of the target's disclosures were not significantly correlated, indicating that disclosure patterns among the elderly were not reciprocal. in addition, having a relationship characterized by reciprocal intimate disclosure was positively, rather than negatively correlated with depression. this indicates that the more depressed an elderly individual is, the more the individual had in objective terms - confidants. one interpretation of this latter finding was advanced, that depressed elderly reciprocate personal information in the form of negative perceptions or complaints.