OBJECTIVE: Questions have been posed about the competence of persons with serious mental illness to consent to participate in clinical research. This study compared competence-related abilities of hospitalized persons with schizophrenia with those of a comparison sample of persons from the community who had never had a psychiatric hospitalization. METHODS: The study participants were administered the MacArthur Competence Assessment Tool for Clinical Research (MacCAT-CR), a structured instrument designed to aid in the assessment of competence to consent to clinical research. The scores of 27 persons who met DSM-IV criteria for schizophrenia who were long-stay patients on a state hospital research ward were compared with those of 24 individuals from the community who were of similar age, gender, race, and socioeconomic status. RESULTS: Significant differences were found between the patients and the community sample on three measures of competence-related abilities: understanding, reasoning, and appreciation. Degree of psychopathology and cognitive functioning were significantly negatively correlated with understanding and appreciation among the patients with schizophrenia. Length of hospitalization was significantly negatively correlated with all measures of decision-making capacities. CONCLUSIONS: The generally poor performance of the long-stay patients with chronic schizophrenia highlights the difficulties this group is likely to encounter in providing consent to research. However, variation across the sample points to the need for individualized assessment and for validated techniques for facilitating decision making in the face of decisional impairments.