Two studies explored how counselor and client agreement on the therapy alliance, at the beginning of treatment, influenced early session evaluations and symptom change. Unlike prior studies that operationalized alliance convergence as either a profile similarity correlation or a difference score, the present study used polynomial regression and response surface analysis to examine agreement. Study 1 explored the impact of working alliance congruence on session depth and smoothness at the 3rd session of treatment with 36 client-counselor dyads. Results revealed that session smoothness was greater when clients' and therapists' perceptions of the working alliance were in agreement and high compared with when they were in agreement and low. In addition, clients rated sessions less smooth when their ratings of the alliance were lower than their therapists' ratings of the alliance, and they rated sessions as more smooth when their ratings of the alliance were higher than their therapists' ratings of the alliance. The authors did not find a significant relationship with session depth. In Study 2, the authors explored the impact of working alliance congruence, at the 3rd session of therapy, on symptom change for 63 client-counselor dyads. Results revealed that as the therapist and client have more positive agreement on the perceived alliance at the beginning of the treatment, there is greater symptom change. The authors also found that the consequences of alliance disagreement are the same regardless of who rated the alliance higher than the other. Implications and recommendations for future research are discussed.