The outcome of clients who saw one of four "expert" professional group therapists selected by peer nomination or four "natural helper" nonprofessional nominated by students is contrasted in a 15-session psychotherapy group. Process measures tapping specific group and "common factors" were drawn from sessions 3, 8, and 14; outcome was assessed at pre, mid, posttreatment, and a 6-month follow-up. Results were examined by leader condition (professional vs. nonprofessional therapists) and time (group development). Virtually no reliable differences were found on measures of outcome primarily because of a floor effect on several measures. Therapist differences on the process measures tapping the "common factors" of therapeutic alliance, client expectancy, and perception of therapists were either nonsignificant or disappeared by the end of treatment. A complex picture of differences on one therapeutic factor (insight), common factor measures and subtle variation in the outcome data suggests a distinct pattern of change, however. Methodological limitations are also addressed including problems inherent in large-scale clinical-trial studies, ethical concerns raised by using nonprofessional leaders, and problems with generalizability, given the absence of significant psychopathology in group members.