The therapeutic alliance predicts dropout from psychological therapy, and ruptures in the therapeutic alliance may also predict dropout, yet there is a dearth of research with adolescents. This study investigated whether markers of rupture-repair in the alliance were indicative of different types of treatment ending in adolescents receiving psychological treatment for depression. Data were from the IMPACT study, a trial investigating the effectiveness of therapies for adolescent depression. Participants were randomly allocated to receive a brief psychosocial intervention, cognitive-behavioral therapy, or short-term psychoanalytic psychotherapy. The sample (N = 35) comprised adolescents who had either completed their treatment (n = 14) or dropped out (n = 21) according to their therapist. Dropout cases were further classified as dissatisfied (n = 14) or got-what-they-needed (n = 7) based on posttherapy interviews with the adolescent and therapist. Selected recordings of therapy sessions were rated using the Rupture Resolution Rating System and Working Alliance Inventory (observer version). Therapeutic alliance and rupture-repair during therapy were similar for completers and got-what-they-needed dropouts, whereas dissatisfied dropouts had poorer therapeutic alliance, more ruptures, ruptures that were frequently unresolved, and greater therapist contribution to ruptures. Qualitative analysis of sessions led to the construction of three categories of therapist contribution to ruptures: therapist minimal response, persisting with a therapeutic activity, and focus on risk. Results suggest that ruptures, especially when unresolved, could be warning signs of disengagement and dropout from therapy. Future research should investigate how ruptures may be effectively identified and resolved in treatment with adolescents. (PsycInfo Database Record (c) 2020 APA, all rights reserved).