OBJECTIVE: The feasibility, reliability, and validity of a new instrument, the MacArthur Competence Assessment Tool-Treatment (MacCAT-T), which was developed for use by clinicians, was tested. The instrument assesses patients' competence to make treatment decisions by examining their capacities in four areas--understanding information relevant to their condition and the recommended treatment, reasoning about the potential risks and benefits of their choices, appreciating the nature of their situation and the consequences of their choices, and expressing a choice. METHOD: The MacCAT-T and instruments to measure symptom severity were administered to 40 patients recently hospitalized with schizophrenia or schizoaffective disorder and 40 matched subjects in the community without mental illness. RESULTS: A high degree of ease of use and interrater reliability was found for the MacCAT-T. Overall, the hospitalized patients performed significantly more poorly than the community subjects on understanding and reasoning, although many patients performed as well as community subjects. Poor performance was related to higher levels of some psychiatric symptoms, such as conceptual disorganization, hallucinations, and disorientation. CONCLUSIONS: The MacCAT-T offers a flexible yet structured method with which caregivers can assess, rate, and report patients' abilities relevant for evaluating competence to consent to treatment.