A recent review of published and unpublished data suggests that the elderly are underrepresented in most self-help groups. Some national groups, such as Alcoholics Anonymous do serve the aged but many others such as widow groups do not. A recent national survey of over 3,000 households conducted by Mellinger and Balter examined the use of psychotropic medication and the help-seeking patterns of troubled respondents. A secondary analysis of the Mellinger-Balter data was conducted to examine patterns of participation in self-help groups by the elderly, the effectiveness of self-help groups, and the differential patterns of participation based on problem or disease to determine what types of groups are most successful in serving the elderly. Critical policy issues raised by these analyses involve the following: (1) the relative value of age-homogeneous groups when compared with age-heterogeneous groups; (2) the best forms of legitimization of self-help groups; and (3) the role of the professional service system. These findings have implications for the different strategies of generating and maintaining self-help groups including professional transfer models, financial aid, and consultation models. A reference list and 15 data tables are appended. (Author/ABL)