The literature indicates that whether or not schizophrenic patients are reported to have significant lateral ventricular enlargement depends on control, and not schizophrenic-group values. This discrepancy does not result from differences in age, the ratio of males to females, the number of control subjects used in each study, or whether control groups are comprised of normal subjects or medical patients. However, medical-patient controls tend to have smaller ventricles than do normal individuals. Thus, we assessed lateral-and third-ventricle size and the degree of cortical atrophy in 30 normal volunteers, 30 medical patients, and 30 chronic schizophrenic patients. The use of a medical control group seemed to result in underestimates of ventricle and sulcal size in the normal population and, therefore, overestimates of these values in schizophrenic groups.