Objective: To estimate the effect of education on the risk of Alzheimer's disease (AD). Methods: 373 patients diagnosed with AD and 559 healthy control individuals without first degree relatives with known dementia, were included in a case-control study (2003-2006). All individuals were genotyped for APOE alleles. Odds ratio (OR) for developing AD was calculated by binary logistic regression, with the number of APOE e4 alleles and educational level as covariates. Analyses were carried out separately for men and women and for different age groups. Results: Carriers of one APOE e4 allele had OR of 4.2, and carriers of two APOE e4 alleles OR of 12.4 for developing AD. When adjusted for the number of APOE e4 alleles, OR for developing AD was significantly reduced in participants with 8-9 years of education compared to those with only 6-7 years, and was reduced further for those with 10-18 years of education. These findings were obtained for all the age groups studied and for both men and women. Conclusions: Education had a consistently protective effect on the risk of developing clinical AD in a dose-dependent manner in both men and women, and in all age groups, also when adjusting for the number of APOE e4 alleles. Male gender was protective, probably at least in part because of a higher educational level. Copyright © 2008 John Wiley & Sons, Ltd.