Background Deficits in the communication and identifying of feelings are usually observed in addiction disorders. These dysfunctions vary according to the type of addiction and are particularly marked for alcoholic subjects. The prevalence of alexithymia evolves in a nearly linear manner according to the severity and the duration of the disorder. As the duration of alcoholism and the quantity of alcohol that is consumed increase, so will the subjects’ scores of alexithymia. In addition, certain authors have observed a decrease in alexithymia in abstinent subjects. Subjects having been abstinent for a long period of time were more alexithymic than those having been abstinent for a shorter period of time. However, other studies failed to confirm these findings and did not observe an increase in the levels of alexithymia in polydrug patients. It seems however that the measures of alexithymia obtained using self evaluation tools (TAS 20) lack in precision due to the fact that, even though the subject is supposed to have a deficit in the identifying and the verbalisation of his emotional states, he is asked to do his own evaluation of the said emotional states. Hence, other tools such as the level of emotional awareness scale (LEAS) offer another approach to the measuring of alexithymia, based on the everyday situations and integrated in a general model of the differentiation of emotional states (or of emotional awareness). Design of study The purpose of the study is to describe the level of emotional differentiation of alcoholic subjects, and to evaluate the development of the emotional processes following alcohol cessation. Our sample consists of 88 subjects distributed in three groups: a group of alcoholic participants who are in the process of quitting alcohol (33 subjects), a group of alcoholic participants that have been abstinent for more than 6 months (20 subjects) and a group of control participants (35 subjects). The prevalence of alexithymia is estimated by the Toronto alexithymia scale (TAS 20) which measures three factors: the difficulty to identify one's feelings (TAS 1), the difficulty to express one's feelings (TAS 2), and thoughts which are directed towards the outside world (TAS 3). The levels of emotional consciousness are estimated by the LEAS which establishes 20 scenarios in order to measure three scores: emotional consciousness “for oneself” (LEAS 1), emotional consciousness “for others” (LEAS 2) and a total score (total LEAS). Results The results show a different level of emotional awareness and alexithymia between the three groups of participants. The alcoholic participants are less conscious of their feelings and have more difficulties identifying and expressing their feelings than the control subjects. With regards to the evolution of the emotional deficit linked to alcohol withdrawal, the abstinent subjects obtain weaker scores of alexithymia than the alcoholic subjects. However, abstinent subjects continue to present more difficulties expressing their feelings than control subjects (TAS 2). On the other hand, abstinent subjects’ levels of emotional awareness are once again comparable to those of control subjects. Finally, the analysis of the correlations does not show any link between the scores of alexithymia and the scores of emotional consciousness, and this for all of the groups studied. Our major result concerns the recovery of the emotional deficit of abstinent subjects (global score of TAS 20 and the score of emotional consciousness), which underlines the importance of the implementation of therapeutic protocols that focus on the emotional awareness of alcoholic subjects.