There is a long-recognized association between alcohol consumption and aggressive behavior. This study was designed to examine aggression in a group of socially well-adapted recovered alcoholics (RA). The question addressed was whether the treatment, together with long-term abstinence from alcohol, could reduce aggression and hostility in RA. A group of male RA (n = 64), who did not meet the DSM-IV criteria for any psychiatric or personality disorder, were recruited to the study from aftercare groups. According to data from their group therapists, they were reliably abstinent for at least 3 years and socially well adapted. The study participants representing the control group (n = 69), diagnosed as being ‘reliable nonalcoholics’ (NA) by the Munich Alcoholism Test, were recruited from general practice. Data were derived from an in-house questionnaire on general characteristics of both groups, and aggressive and hostility traits were assessed using the Buss-Durkee Hostility Inventory (BDHI). The univariate and multivariate between-groups design was used for data analysis. Taking into account the BDHI dimensions of aggression and hostility, the difference between RA and NA groups was statistically significant [Wilks’ lambda (8, 125) = 0.769; p = 0.00004]. There were statistically significant differences in the BDHI scales for indirect aggression, irritability, negativism, suspicion, resentment, and guilt. Both RA and Na groups did not differ significantly in variables that assessed physical and verbal aggression. After a 3-year abstinence, subjects from the RA group displayed signs of hostility and covert aggression.