Introduction: Forensic mental health care is hampered by lack of evidence-based treatments. The Swedish forensic mental health population consists of patients suffering from severe illnesses such as schizophrenia and bipolar disorders, similar to populations in international studies. Illness Management and Recovery (IMR) is an intervention for patients with serious mental illness, based on psychoeducational, cognitive-behavioral and motivational components. The purpose is to strengthen participants’ illness management skills and recovery. Objective: To test effectiveness of IMR within forensic mental health by comparing it to treatment as usual. Method: This is a cluster-randomized controlled trial. Patients in forensic mental health inpatient units are randomized to an active (IMR) or a control condition (treatment as usual). Clustering of patients is based on ward-units where inpatients are admitted. Patients in the active condition receive two group and one individual IMR sessions per week. The treatment phase is estimated to last nine months. Outcomes include illness related disability, illness management skills, sense of recovery, hope, mental health and security related problems. Outcomes are measured at baseline, four months into treatment, at treatment completion and at three months follow-up. Staff experiences of implementing IMR will be explored by a self-report measure and semi-structured interview based on Normalization Process Theory. Ethics and dissemination: The study is approved by the Swedish Ethical Review Authority (Registration No. 2020–02046). Participation will be voluntary based on written informed consent. Results will be disseminated through peer-reviewed articles and conferences. The study is registered in the US registry of clinical trials (NCT04695132).