An understanding of why people smoke cigarettes can have an important impact on smoking prevention and cessation. People smoke cigarettes to maintain nicotine levels in the body, and nicotine has been implicated in the stimulation of brain reward mechanisms via central neuronal dopaminergic pathways. In this study, we evaluated the association of smoking and smoking cessation with a dopamine D4 receptor 48-bp variable nucleotide tandem repeat polymorphism in which the seven-repeat allele (D4.7) reduces dopamine affinity. Smokers (n = 283) and nonsmokers (n = 192) were recruited through local media for a case-control study of smoking. After giving informed consent and answering a behavioral questionnaire, smokers underwent a single minimal-contact session of smoking cessation counseling and then were followed for up to 1 year. The frequency of the dopamine D4 receptor genetic polymorphism using PCR was determined, and individuals were classified by the number of repeat alleles (two to five repeats as S and six to eight repeats as L). Persons with those genotypes including only S alleles (homozygote S/S) were compared with those with at least one L allele (heterozygote S/L and homozygote L/L). Chi2 tests of association, Fisher's exact test, and Student's t test were used. Ps were two-tailed. The data show that African-Americans (n = 72) who had at least one L allele had a higher risk of smoking (odds ratio, 7.7; 95% confidence interval, 1.5-39.9; P = 0.006), shorter time to the first cigarette in the morning (P = 0.03), and earlier age at smoking initiation (P = 0.09) compared with homozygote S/S genotypes. After smoking cessation counseling, none of the African-American smokers with an L allele were abstinent at 2 months, compared with 35% of the smokers who were homozygote S/S (P = 0.02). The analysis of Caucasians (n = 403) did not suggest a similar smoking risk for the D4 genotypes (odds ratio, 1.0; 95% confidence interval, 0.6-1.6; P = 0.90), or smoking cessation (P = 0.75). Although the number of African-Americans is small, this study is consistent with the hypothesis that the L alleles increase the risk of smoking because these individuals are prone to use nicotine to stimulate synaptic dopamine transmission. If replicated, the data indicate that a single minimal-contact session of cessation counseling, similar to what is typically provided in primary care physician offices, is ineffective in African-American smokers who have at least one L allele. The finding of an effect for these polymorphic loci in African-Americans, but not Caucasians, suggests that the variable nucleotide tandem repeat studied here is a marker for another polymorphic site in African-Americans, but not in Caucasians.