Genome-wide studies of linkage disequilibrium have revealed that most of the human genome can be divided into blocks of varying length within which marker to marker linkage disequilibrium is very high. Each haplotype block is separated by recombination sites. There are usually only 3-6 haplotypes with greater than 5% frequency within each block. These haplotypes reflect descent from a single, ancient ancestral chromosome. The main advantage of haplotype methods for linkage and association studies is that these common haplotypes capture most of the information on genetic variation within these regions and the haplotypes can be identified using only a small number of SNPs, usually 3 to 8. Thus haplotype-based case-control studies can detect associations with disease or behavior without having to find and test every single variant in the region. We have genotyped two ethnically diverse population isolates, approximately 500 Finnish Caucasians and 400 Plains American Indians, and one admixed population, 900 African Americans, for 130 candidate genes for alcoholism and anxiety, including the chromosome 4 cluster of GABAA receptor genes that are predominantly expressed in the brain reward circuitry. In both Caucasians and Plains Indians we found GABRG1 haplotype linkage to alcoholism. Moreover, long-distance ancestral haplotypes spanning GABRG1 and GABRA2 were associated with alcoholism;this association was determined by GABRG1 (Enoch et al, 2009). The African Americans had four common GABRA2 haplotypes within the distal haplotype block: two that correspond to the Caucasian and Asian yin-yang haplotypes and two not found in other ethnic groups. One of the unique haplotypes predicted heroin addiction whereas the other haplotype was more common in controls. Moreover, an intron 2 SNP rs11503014, not located in any haplotype block and potentially implicated in exon splicing, was independently associated with addiction, specifically heroin addiction (Enoch et al,in press, Biol Psychiatry). The neurotransmitter serotonin (5-HT) has a profound influence on mood, and CNS 5-HT variation has been implicated in alcohol and drug dependence. The African American men showed a haplotype association with alcohol dependence that was driven by a gain of function SNP, rs1176744, Tyr129Ser. Moreover, we showed that although the HTR3B rs1176744 12/22 gain of function genotype predicted alcohol dependence and the low activity triallelic serotonin transporter promoter variant 5-HTTLPR predicted cocaine dependence, the two HTR3B and 5-HTTLPR genotypes combined synergistically in their effects on addiction to both alcohol and cocaine. The mechanism underlying this synergism might be that increased synaptic 5-HT plus increased 5-HT3 receptor responsiveness to 5-HT might result in alterations in dopamine release in the reward pathway that increase vulnerability to addiction (Enoch et al, in preparation). We investigated whether variation in alcohol metabolism genes might alter the risk for alcohol dependence in non-East Asian populations. We performed comprehensive haplotype association analyses based on 64 haplotype tagging SNPs across ALDH1A1, ALDH2 and all seven genes in the ADH cluster in Finnish Caucasians, African Americans, Plains American Indians and Southwestern American Indians. Our study found several results that are of interest, particularly for ALDH1A1 and ADH4 (Liu et al, submitted). We have previously shown that the functional COMT Val158Met polymorphism is associated with trait anxiety in women (Enoch et al, 2003). We have recently shown that individuals with anxiety disorders comorbid with alcoholism and/or depression have the greatest proportion of COMT Met158 and BDNF Met66 alleles compared with individuals with pure anxiety, alcoholism, major depression or no diagnosis (Enoch et al, 2008). Furthemore, in Plains Indians, the COMT Met158 allele is associated with better long-term memory, working memory and attention (Enoch et al, 2009).