Objective Despite several studies that have highlighted the harmful effects of alcohol consumption on cognitive functions it remains unclear whether certain brain areas are more sensitive than others are or whether alcohol causes widespread cognitive deficit. Moreover, the role of continued abstinence has yet to be clarified regarding the quality of recovery on the different cognitive domains. The aim of this 1-year longitudinal study was to evaluate the recovery of cognitive deficits in the medium (6 months) and long term (12 months) after the interruption of drinking. Methods Forty-one alcohol-dependent patients were recruited from two outpatient treatment facilities and cognitive functions were compared on a control group of forty healthy controls. The patients were then re-assessed at 6 and 12 months. Changes in neuropsychological measures were evaluated with repeated measures analysis of variance (ANOVA). We also compared 1-year follow-up scores with control data (unpaired t tests) to identify tests on which significant differences persisted. Results Patients performed significantly worse than controls in all cognitive domains investigated and this cognitive impairment was evident in recently abstinent patients. A year of abstinence resulted in a significant improvement in all cognitive domains assessed after detoxification from alcohol. After year 1, alcoholic subjects had returned to normal levels compared to healthy controls on all domains except for general non-verbal intelligence, verbal memory and some visuospatial skills. Conclusion Our results support the hypothesis of widespread impairment resulting from alcohol consumption. The recovery of cognitive functions is not homogeneous during prolonged abstinence.