The involvement of dopamine in the etiology of schizophrenia is suggested by a number of neurobiological and pharmacological data, the dopamine D3 receptor (DRD3) being selectively expressed in brain regions which may be specifically involved in the risk for schizophrenia. The gene coding for DRD3 has thus been extensively analyzed. Since the initial report providing substantial evidence for an association of homozygosity of either allele of the gene coding for DRD3 (BalI polymorphisms) with schizophrenia, a flurry of replicating studies has appeared, which have been split into confirmations and nonreplications in North European Caucasian, Mediterranean, Asian, American, and African populations. The involvement of DRD3 polymorphisms thus remains questionable, particularly as no linkage studies have favored a nonrandom segregation of DRD3 alleles and schizophrenia. We performed a metaanalysis from 29 independent samples, from 24 different association studies so far published, allowing the examination of 2,619 schizophrenic patients and 2,517 controls. No significant differences of genotype counts were noted between patients and controls for the whole sample, considering frequency of any genotype. Starting from the high variability of the genotypes in different geographical areas, the impact of ethnic heterogeneity was taken into account. When the studies were reorganized in five groups according to geographical origin of samples, both homozygosity and 1-1 genotype revealed significant heterogeneity (P < 0.05). We specifically found an excess of homozygosity and 1-1 genotype in schizophrenic patients only in the African and Caucasian groups (P < 0.05). The present analysis suggests a small but significant effect of DRD3 in the susceptibility to schizophrenia, at least in Caucasians.