This study sought to determine the therapeutic effectiveness of client-therapist dyads in a residential treatment center for emotionally disturbed adolescents. The theories of George Kelly's personal construct psychology were utilized in assessing the dyadic relationship. The four elements investigated were organizational similarity, understanding, organizational congruency and predominant selves. The sample consisted of 140 dyads comprised of 10 adolescent boys and girls and 14 therapeutic staff of a residential treatment center in the southwest United States. Responses to Kelly's Role Construct Repertory Test were compared to four relational factors—parental/respect, identity, problem-solving, and sexual/affection—and two rating scales of client-therapist preference and ratings of therapeutic effectiveness. Contrary to expectations, as content similarity among dyads composed of clients and staff increased, there was not an increase in functional aspects of the therapy relationship. Possible mitigating factors may have been level of client disturbance and/or methodological issues relating to how organizational similarity was determined. Dyadic understanding was not found to be related to perceptions of the therapy relationship. This may be a function of adolescent of adolescent clients' need for independence and resistance to adult understanding and control. Therapy dyads with a moderate level of lateral or vertical organizational congruence were not found to be curvilinearly related to functional aspects of the therapy relationship. However, a weak linear relationship regarding client perceptions of the therapy relationship was noted on four measures. Several methodological recommendations related to the instruments used to determine therapeutic effectiveness and the means of eliciting personal constructs on the REP test.