A comparative analysis is presented of adolescent participation in three-person family problem-solving groups, involving five families with an adolescent diabetic and three families with an adolescent with no diagnosed illness. Observation and interaction analysis of family problem-solving in a laboratory setting were used, as families were asked to discuss written vignettes describing three problem-solving situations. Problem-solving behaviors were then coded into seven summary codes, and the problem-solving process was organized into four phases (introduction, alternative generation, evaluation, and decision). Results indicated that the adolescents in both groups were actively involved in problem-solving, but in different ways. Diabetic families had a more balanced distribution of participation among the mothers, fathers, and teens. No difference was found between the two family groups in terms of total family participation by summary code; the problem-solvers used most of their thought units on identifying the problem and the rest in the following order: evaluating the problem, resolving it, offering alternatives, and least of all, offering decisions. Non-diabetic teens, however, contributed a relatively high percentage of decisions and alternatives and a relatively low percentage of identifications. In an analysis of participation patterns by phases, non-diabetic teens again sent proportionately more decisions than did the other participants. (JDD)