Despite widespread clinical, theoretical, and empirical support for the importance of alliance ruptures, little is known about the underlying biological level at times of rupture. The overarching goal of the present study was to investigate dyadic patterns of in-session oxytocin (OT) change between patients and therapists (e.g., patient's OT increases more than therapist's OT) as markers of withdrawal ruptures. Hypothesis 1 construed that OT incongruence (e.g., larger patient increase in OT in comparison to their therapist OT increase) will mark the occurrence of withdrawal ruptures. Hypothesis 2 construed that this effect of OT incongruence will be more pronounced when anxious attachment orientation is low. Surface analysis was conducted on 628 saliva samples that were gathered before and after therapeutic sessions of 53 patient-therapist dyads enrolled in a randomized control trial treating major depression. Only Hypothesis 2 received empirical support, meaning it was only when anxious attachment orientation was low that there were significantly more withdrawal ruptures when the patient's OT increase was higher than their therapist's OT increase. This is consistent with the literature, suggesting that in times of withdrawal ruptures, the patient and therapist are in an incongruent state. Findings suggest that this incongruence is mirrored at the biological level only when anxious attachment orientation is low. Results shed light on what may be happening between patients and therapists on a biological level during a withdrawal rupture. (PsycInfo Database Record (c) 2023 APA, all rights reserved).