Several case-control studies have reported an association between schizophrenia and the epsilon4 allele of Apolipoprotein E gene. The results have been equivocal. This meta-analysis has evaluated the collective evidence for an association between the epsilon4 allele of Apolipoprotein E gene and schizophrenia. We analyzed published data sequentially first considering epsilon4 allele itself, and then epsilon4 carrier status as risk factors for schizophrenia using a sample of 17 population-based case-control studies, of which 6 were from Asian and 11 from Caucasian populations. The pooled odds ratios from the Caucasian populations showed a modest association with risk of schizophrenia for epsilon4 allele and epsilon4 carrier genotype. No other alleles or genotypes were significant in either Asian or Caucasian populations when analysed separately or combined, although the sample size had over 80% power to detect a significant odds ratio of 1.9 in Asian-population studies and 1.6 in Caucasian-population studies. After allowing for sensitivity analysis of the studies and assessment of publication bias, we conclude that the epsilon4 allele of Apolipoprotein E does not play a major role in risk of schizophrenia in Caucasian populations. Since significant heterogeneity was present among the 6 Asian populations reported to date, further studies using larger sample sizes are required.