BACKGROUND: Command hallucinations are among the most distressing, high risk and treatment resistant symptoms for people with psychosis; however, currently, there are no evidence-based treatment options available for this group. A cognitive therapy grounded in the principles of the Social Rank Theory, is being evaluated in terms of its effectiveness in reducing harmful compliance with command hallucinations.METHODS/DESIGN: This is a single blind, intention-to-treat, multi-centre, randomized controlled trial comparing Cognitive Therapy for Command Hallucinations + Treatment as Usual with Treatment as Usual alone. Eligible participants have to fulfil the following inclusion criteria: i) ≥16 years; ii) ICD-10 diagnosis of schizophrenia or related disorder; iii) command hallucinations for at least 6 months leading to risk of harm to self or others. Following the completion of baseline assessments, eligible participants will be randomly allocated to either the Cognitive Therapy for Command Hallucinations + Treatment as Usual group or the Treatment as Usual group. Outcome will be assessed at 9 and 18 months post randomization with assessors blind to treatment allocation. The primary outcome is compliance behaviour and secondary outcomes include beliefs about voices' power, distress, psychotic symptoms together with a health economic evaluation. Qualitative interviews with services users will explore the acceptability of Cognitive Therapy for Command Hallucinations.DISCUSSION: Cognitive behaviour therapy is recommended for people with psychosis; however, its focus and evaluation has primarily revolved around the reduction of psychotic symptoms. In this trial, however, the focus of the cognitive behavioural intervention is on individuals' appraisals, behaviour and affect and not necessarily symptoms; this is also reflected in the outcome measures used. If successful, the results will mark a significant breakthrough in the evidence base for service users and clinicians and will provide a treatment option for this group where none currently exist. The trial will open the way for further breakthrough work with the 'high risk' population of individuals with psychosis, which we would intend to pursue.