Numerous studies have tested for associations between an intronic polymorphism (rs165932) of presenilin-1 (PS-1) gene and the risk of Alzheimer's disease (AD), but results have been conflicting. To throw light on this issue, we investigate the possible involvement of PS-1 genotype in a case-control study based on a relatively stable population in Spain and a meta-analysis of published studies. An examination was conducted of 85 patients with probable or possible AD, along with controls from the same community, by using an chi(2) test for homogeneity and a binary logistic regression model. For comparison purposes, a meta-analysis of data from all available published studies was assessed. In our patients, homozygosity of the allele 2 in the PS-1 gene increased for late-onset AD (OR 2.38, 95% CI 1.07-5.29, P<0.05). The presence of at least one allele of apoE was also associated with AD (OR 4.01, 95% CI 1.93-8.34, p<0.05). The regression model showed that, overall, the presence of the apoE epsilon 4 allele and the PS-1 2/2 genotype were independent factors for the development of AD in our sample. In our genotype-based meta-analysis, the PS-1 2/2 genotype was probably related with AD for the European sub-group (fixed effects model, OR 1.19, 95% CI 1.02-1.37, p<0.05), but there are many confusing factors between different studies. Presenilin-1 2/2 genotype is a risk factor for late onset Alzheimer disease in the Spanish population, and probably, for Europeans.