The ability to generate antisaccades (eye movements deliberately made in the direction opposite to that of a visual stimulus) may be used to assess central nervous system function in a variety of neurologic and psychiatric disorders. However, the usefulness of this paradigm in clinical practice is limited by the need for an oculographic laboratory. We describe a clinical version of such an antisaccadic task and present normative data from 332 subjects. We also examined clinical antisaccades and cognitive performance in 30 patients with Alzheimer's disease, five patients with Huntington's disease, and 12 patients with pseudodementia. In Alzheimer's disease, error rates in the clinical antisaccadic test correlated well with those from a laboratory-based antisaccadic task measured on the same day by infrared oculography, confirming that the clinical antisaccadic test is a valid analog of the more sophisticated laboratory paradigms. Clinical antisaccadic error rates correlated strongly with the severity of dementia in Alzheimer's disease, and correlations with cognitive performance suggested that the clinical antisaccadic test may have some specificity for frontal lobe dysfunction. Patients with pseudodementia had normal clinical antisaccadic error rates, and the test may therefore be of use in differentiating dementia from pseudodementia. This clinical antisaccadic test provides a simple, reliable, and inexpensive quantitative clinical tool that is of value in the assessment of disturbances of higher cortical function.