Abstract Background Process-based therapy (PBT) is a new framework to intervention planning, based on the use of ecological momentary assessment (EMA) data and dynamic and idiographic network analyses. Support for its applicability has been reported from a single-case studies. Here, we examine the feasibility and effectiveness of PBT in a larger clinical sample. We have translated a training manual of PBT and modified for delivery of CBT in mental health service. The aim of this study is to test the relative efficacy of PBT compared to traditional CBT delivered in routine practice (r-CBT) for difficult-to-treat mood and anxiety disorders. Methods The study is a randomized controlled trial (RCT) of PBT vs r-CBT for difficult-to-treat unipolar depression and anxiety disorders. In total, 80 patients are recruited at an outpatient clinic and included in two intervention arms. Primary outcome is emotional distress; secondary outcomes include psychological well-being and quality of life, adaptive behavior, psychological flexibility, and reflective functioning. Assessments of outcome variables are conducted before and after therapy and at 6 months follow-up. Weekly patient-rated outcomes are collected for every session to investigate process of change. Outcome assessors, blind to treatment allocation, will perform the observer-based symptom ratings, and adherence with manual will be monitored using self-report. Discussion The current study will be the first RCT of PBT in a health care setting. The planned moderator and mediator analyses will clarify the mechanisms of change in psychotherapy and the association between personalized assessment based on dynamic network analysis and treatment effect. Trial registration ClinicalTrials.gov NCT06517589. Registered 24 July 2024.