The WMS-R represents a significant improvement over the original version of the WMS. Patterns of performance on the five WMS-R indices successfully differentiated patients with a "cortical" (that is, DAT) from patients with a "subcortical" (that is, HD) dementia. The differences between ACI and GMI also indicated that DAT patients' attention and concentration were better preserved than those of equally demented HD patients. Analyses of the differences between GMI and DMI indicated that rapid forgetting was more apparent for DAT than for HD patients in the early stages of these disorders. When savings scores for the Logical Memory and Visual Reproduction tests were used to examine retention over a 30-minute period, both mildly and moderately demented DAT patients demonstrated more rapid forgetting than did HD patients with similar levels of cognitive decline. Discriminant analyses performed with these indices of forgetting showed 95 per cent accurate classification of DAT and elderly controls, and 79 per cent correct classification of DAT and HD patients. Assessment of the patients' tendencies to make prior-item intrusion errors on the Visual Reproduction tests of the WMS and WMS-R yielded significant differences between patient groups and the two versions of the test. Patients with DAT made significantly more prior-figure intrusion errors than did HD patients on both versions of the VRT, but the original WMS form of this figural memory test was more likely to elicit such errors than was the revised version. Despite many improvements in the revised scale, a number of limitations still exist including standardization and a lack of norms for elderly individuals, as well as failures to include tests of verbal recognition and adequate measures of nonverbal memory.