A 3-year longitudinal study was conducted with 50 Alzheimer's disease patients and 45 elderly control subjects. All study participants received an extensive evaluation that included brain CT at baseline and follow-up. Quantitation of ventricular size, using both linear and volume methods, revealed highly significant cross-sectional and longitudinal differences between the Alzheimer patients and control subjects. Specifically, the annual rate of change in ventricular volume was approximately 9% in the Alzheimer patients and approximately 2% in the controls. The presence of age-related white matter lesions had no effect on the clinical course of the patients or on the changes in ventricular size. Among the Alzheimer patients, the rate of clinical decline was strongly related to the rate of change in ventricular size. Baseline ventricular measurements were of no value in predicting the subsequent rate of clinical deterioration or ventricular enlargement. The results suggest that changes in ventricular size closely reflect the clinical changes in Alzheimer patients.