Apolipoprotein E (ApoE) genotyping was conducted in sporadic Alzheimer's disease (AD, n = 91) as well as in other dementing disorders including Parkinson's disease (PD, n = 73), autopsy-confirmed diffuse Lewy body disease (DLBD, n = 16), progressive supranuclear palsy (n = 13), vascular dementia (n = 55), alcoholic dementia (n =25) and normal control subjects (n = 77). ApoE epsilon 4 allele frequency was significantly higher in AD (33.5%, p < 0.001), DLBD (40.6%, p < 0.001) and demented PD (29.4%, p < 0.05) compared to that in normal controls (11.7%). The association of the ApoE epsilon 4 allele with AD was more pronounced in early-onset AD (46.4%) than in late-onset AD (27.8%). 46% of the AD individuals developed AD without association to ApoE epsilon 4, and epsilon 4 homozygotes were found not only in AD, but also in many of other dementing disorders. These results suggest that ApoE genotyping cannot provide certainty about the presence of absence of AD, and that it should be used as an adjunct to other diagnostic tests for AD. On the other hand, cerebrospinal fluid (CSF) tau levels were significantly elevated (p < 0.0001) in AD (78.0 +/- 44.2 pg/ml) compared to those in normal controls (10.6 +/- 8.6 pg/ml). The specificity and the sensitivity of distinguishing AD from normal controls was 95.0 and 91.2%, respectively. Elevated CSF-tau levels were also detected in some patients with acute neurological diseases including meningoencephalitis, Creutzfeld-Jacob disease, normal pressure hydrocephalus and vitamin B12 deficiency encephalopathy. Increased CSF-tau levels in AD were found regardless of the age at onset, clinical stage, ApoE genotype, alpha 1-antichymotrypsin genotype, and presenilin-1 genotype. The CSF-tau levels continued to be abnormal during the progression of AD. These results suggest that CSF-tau serves as an unequivocal and reliable biological marker to aid in the clinical diagnosis of AD.