Objective: The goal was to understand both therapist and patient perspectives on alliance and session progress for women in treatment for gynecological cancer. We used a longitudinal version of the one-with-many design to partition variation in alliance and progress ratings into therapist, patient/dyad, and time-specific components. We also evaluated therapist and patient characteristics that predict alliance and session progress. Method: Two hundred and three women and their therapists completed measures of alliance and session progress across a 6-session course of treatment. Participants also completed preintervention measures of self-esteem, depression, cancer-specific distress, emotional expressivity, and use of protective buffering. Results: Patients reported higher alliance and greater progress than did therapists. When therapists reported particularly strong alliance with particular patients, those patients concurred. More experienced therapists reported higher alliances and more progress but their patients did not agree. Patients who began treatment in more difficult psychosocial circumstances tended to have less positive session outcomes on average but evidenced more improvement across therapy sessions. Conclusions: Patients rated their alliance and progress more positively than did their therapists, although there was substantial relative agreement between therapists and patients. Alliance and progress improved over time, particularly among patients who evidenced higher levels of distress and poorer physical functioning. More experienced therapists were more confident in their abilities but their patients did not share this perception. (Contains 5 tables and 3 footnotes.)