BACKGROUND: Although people with schizophrenia display impaired abilities for consent, it is not known how much impairment constitutes incapacity. AIMS: To assess a method for determining the categorical capacity status of potential participants in schizophrenia research. METHOD: Expert-judgement validation of capacity thresholds on the sub-scales of the MacArthur Competence Assessment Tool-Clinical Research (MacCAT-CR) was evaluated using receiver operating characteristic (ROC) analysis in 91 people with severe mental illness and 40 controls. RESULTS: The ROC areas under the curve for the understanding, appreciation and reasoning sub-scales of the MacCAT-CR were 0.94 (95% CI 0.88-0.99), 0.85 (95% CI 0.76-0.94) and 0.80 (95% CI 0.70-0.90). These findings yielded negative and positive predictive values of incapacity that can guide the practice of investigators and research ethics committees. CONCLUSIONS: By performing such validation studies for a few categories of research with varying risks and benefits, it might be possible to create evidence-based capacity determination guidelines for most schizophrenia research.