Three investigators have applied different histopathologic methods (modified Bielschowsky silver methods, Congo red-gallocyanin) to differentiate Alzheimer's disease (AD) (n = 7 subjects; four with very mild dementia and three with moderate to advanced dementia) neuropathology from brain changes associated with aging in three nondemented individuals who had been evaluated using a validated dementia severity staging instrument [Washington University Clinical Dementia Rating (CDR)] generally within a year of death. The presence of elevated numbers of neocortical (frontal and temporal) diffuse, mature, and total senile plaques (SP) was strongly correlated with the presence of clinical AD but did not equate with CDR dementia severity. Neocortical neurofibrillary tangle (NFT) density as well as hippocampal NFT and SP density in this small series did not differentiate statistically between AD and controls. NFT density appeared to correlate with CDR better than SP density. Quantitative histopathologic assessment of AD markers in only a few brain regions can accurately predict the presence of clinical AD, including the very mild form of the disease. This is especially true for SP in the neocortex.