We assessed a set of biological (HDL, LDL, SGOT, SGPT, GGT, HTc, Hb and T levels) and psychometric variables (investigated through HAM-D, HAM-A, GAS, Liebowitz Social Anxiety Scale, Mark & Mathews Scale, Leyton scale, and Pilowski scale) in a sample of 64 alcohol dependent patients, at baseline and after a detoxification treatment. Moreover, we recruited 47 non-consanguineous relatives who did not suffer alcohol related disorders and underwent the same tests. In both groups we genotyped 11 genetic variations (rs1800587; rs3087258; rs1799724; 5-HTTLPR; rs1386493; rs1386494; rs1487275; rs1843809; rs4570625; rs2129575; rs6313) located in genes whose impact on alcohol related behaviors and disorders has been hypothesized (IL1A, IL1B, TNF, 5-HTTLPR, TPH2 and HTR2A). We analyzed the epistasis of these genetic variations upon the biological and psychological dimensions in the cases and their relatives. Further on, we analyzed the effects of the combined genetic variations on the short - term detoxification treatment efficacy. Finally, being the only not yet investigated variation within this sample, we analyzed the impact of the rs6313 alone on baseline assessment and treatment efficacy. We detected the following results: the couple rs6313 + rs2129575 affected the Leyton -Trait at admission (p = 0.01) (obsessive-compulsive trait), whilst rs1800587 + 5-HTTLPR impacted the Pilowski test at admission (p = 0.01) (hypochondriac symptoms). These results did not survive Bonferroni correction (p < or = 0.004). This lack of association may depend on the incomplete gene coverage or on the small sample size which limited the power of the study. On the other hand, it may reflect a substantial absence of relevance of the genotype variants toward the alcohol related investigated dimensions. Nonetheless, the marginal significance we detected could witness an informative correlation worth investigating in larger samples.