In this study, we aimed to explore interpersonal distress as a moderator of the effects of supportive treatment (ST) versus supportive expressive treatment (SET) on the therapeutic alliance and outcome in patients with depression. We drew on data from a randomized controlled trial where 100 patients were randomized to receive 16 sessions of either ST or SET. Patients completed a measure of interpersonal problems at baseline. Therapy outcome and the strength of the therapeutic alliance were measured by patient self-report, before and after each session, respectively. We conducted interactive three-level mixed effect models, accounting for the nested structure of the data, with repeated measures nested within patients, nested within therapists. The results showed significant interactive effects of treatment by baseline interpersonal problems on both the evolution of outcome and the alliance. Patients with greater interpersonal problems had a greater reduction in depressive severity and greater improvements in the alliance when treated with SET versus ST. If further replicated, the results of this study suggest that therapists of patients with depression would benefit from evaluating their interpersonal distress at baseline and incorporating this evaluation in their decision-making process regarding the most suitable intervention strategy for each individual patient. To conclude, the results of this study might enhance evidence-based treatment personalization in patients with depression using a parsimonious personalization approach. (PsycInfo Database Record (c) 2025 APA, all rights reserved).