Informed by a trans-theoretical model of the therapeutic alliance in conjoint family therapy, this exploratory study was the initial stage in a task analysis of critical shifts in adolescent engagement. Specifically, we compared sessions in which a resistant adolescent either did or did not shift from negative to positive engagement during the session. Two successful and two unsuccessful change events were selected from an archival data set based on SOFTA ratings of the therapeutic alliance. The results suggested that one parent element (support) and five therapist elements (structuring therapeutic conversations, fostering autonomy, building systemic awareness, rolling with resistance and understanding the adolescent's subjective experience) seemed critical for successfully facilitating adolescent engagement. The qualitative results were informed by the adolescent's self-reported target complaints pretreatment, which suggested varying reasons for the teenagers’ active or passive disengagement. Implications for practice and recommendations for future research are offered to continue this line of inquiry.