Seventeen patients, treated for Wilson's disease (WD), underwent a set of neuropsychological tests and were compared with a closely matched control group. There were clear differences between the groups (chi 2-test, p less than 0.0001). Wilson patients with only hepatic involvement, however, did not at all differ from their controls. Wilson patients with neuropsychiatric signs differed from controls on a reasoning test (p = 0.0016), and the entire WD group differed on a perceptual speed task (p = 0.0025). Compared to normal test values, however, the patients' group means were all within plus or minus one standard deviation from the normal mean. Special testing procedures and construction of the test battery excluded a factor of motor deficits as a major cause for the differences. The neuropsychological findings are viewed in relation to other findings in patients with motor disorders and predominantly subcortical lesion sites. Wilson's disease may be a dementing condition, but not when treated adequately.