Clock drawing has recently been shown to lie useful in differentiating Alzieimer's disease patients from normal controls. Our procedure for clock drawing differed from other published reports in that a copy condition was employed and patients were asked to set clock hands to read "ten after eleven". We found both clock drawing procedures to be correlated with tests related to executive and visuospatial functioning. In both conditions, nondemented controls performed significantly better than demented patients. In the command condition there was no difference between Alzheimer patients and patients with cerebrovascular dementia. In the copy condition, patients with cerebrovascular dementia performed significantly worse than Alzheimer patients. The inclusion of a copy condition appears to greatly expand the utility of this test. Although our scoring system did not differentiate between various dementing disorders in the command condition, if clock drawing is used as a screening instrument, lack of improvement in the copy condition in comparison to the command condition may be a sign of a vascular involvement.