We found strong associations between ADH4 gene variation and alcohol and drug dependence by the Hardy-Weinberg Disequilibrium (HWD) test and case-control association analysis in an initial study. The present study aimed to confirm these findings by controlling for population stratification and admixture effects to which the HWD test and case-control association methods may be vulnerable. In addition to 365 unrelated healthy controls and 560 unrelated cases in the initial study, we evaluated 104 small nuclear families with affected offspring who had diagnoses of alcohol and/or drug dependence. Seven single nucleotide polymorphisms (SNPs) at the ADH4 gene locus were genotyped in all subjects, and 38 unlinked ancestry-informative markers were also genotyped in unrelated cases and controls. Structured association analysis demonstrated that the genotypes of six ADH4 markers were associated with alcohol dependence, and all seven ADH4 markers were associated with drug dependence (P=10-0.047). Logistic regression analysis showed that: (i) the genotypes of SNP2 (rs1042363) were significantly associated with alcohol dependence and drug dependence (mainly cocaine dependence), and the genotypes of SNP3 (rs1126671) were also significantly associated with alcohol dependence and (ii) one seven-variant haplotype and one diplotype were significantly associated with alcohol dependence and other seven-variant diplotypes were significantly associated with drug dependence (including cocaine and opioid dependence). Transmission disequilibrium test, haplotype-based haplotype relative risk and genotype-based haplotype relative risk analyses all confirmed the association of the ADH4 markers with alcohol dependence and drug dependence. Using rigorous study designs that account for possible population stratification, these findings confirm and extend our original observations indicating that variation at ADH4 predisposes to alcohol and drug dependence.