We performed a meta-analysis of over 30 case-control studies of association between schizophrenia and a bi-allelic, Bal I polymorphism in exon 1 of the dopamine D3 receptor gene. We observed a significant excess of both forms of homozygote in patients (P = 0.0009, odds ratio (OR) = 1.21, 95% Confidence Interval (CI) = 1.07-1.35) in the combined sample of 5351 individuals. No significant heterogeneity was detected between samples and the effects did not appear to be the product of publishing bias. In addition we undertook an independent, familybased association study of this polymorphism in 57 parent/proband trios, taken from unrelated European multiplex families segregating schizophrenia. A transmission disequilibrium test (TDT) showed a significant excess of homozygotes in schizophrenic patients (P = 0.004, odds ratio (OR) = 2.7, 95% CI = 1.35-5.86). Although no significant allelic association was observed, a significant association was detected with the 1-1 genotype alone (P = 0.02, OR = 2.32, 95% CI = 1.13-4.99). In addition when the results of the family-based association study were included in the meta-analysis, the homozygosity effect increased in significance (P = 0.0002, OR = 1.23, 95% CI = 1.09-1.38). The results of the meta-analysis and family-based association study provide independent support for a relationship between schizophrenia and homozygosity at the Bal I polymorphism of the D3 receptor gene, or between a locus in linkage disequilibrium with it.