Subjects were followed in the Longitudinal Core Study of the Baltimore Huntington's Disease Center and given annual neurological, cognitive, and psychiatric examinations. Postmortem neuropathological grade was assigned using the system of Vonsattel and colleagues. We examined the correlations between the neuropathological grade and scores on the Quantified Neurological Examination (QNE) and its chorea and motor impairment subscales, the Mini-Mental State Examination (MMSE), the HD Activities of Daily Living (ADL) Scale, and a number of demographic variables (CAG number, age of onset, age at death, disease duration) in 100 subjects who had been examined within 1,000 days of death. All measures showed significant correlation with Vonsattel score except for chorea. The strongest effect was that of motor impairment score (r(2) = 0.351, p < 0.0001). In a stepwise correlation of clinical variables, motor impairment remained significant. The largest effect for a demographic variable was for age of onset (r(2) = 0.226, p < 0.0001) A partial correlation was significant between CAG number and Vonsattel grade, controlling for age at death or disease duration. Motor impairment appears to be a good clinical measure of neuronal cell loss, at least late in the course of HD and therefore may prove useful in observational and treatment studies.