Objective: The investigators hypothesis that VRET associated to cognitive behaviors therapy (CBT) will be more effective than CBT alone on the reduction in the cumulative number of standard drinks (vs) of alcohol intakes at 8 months after inclusion. The second objective is the assess its effectiveness on reported craving also at 8 months. Method: The investigators plane to recruit 156 subjects aged 18 and over, with AUD according to DSM V criteria, abstinent for at least 15 days. Non-inclusion criteria are: AUD relapse, pregnancy or breastfeeding (for women), decompensated comorbid mental disorder, severe cognitive impairment, epilepsy or history of photo paroxysmal EEG responses, balance disorders, recent stroke less than 3 months old, current nausea/vomiting, claustrophobia, severe visual impairment, and medium or high myopia (beyond -3.5 diopters). The study recruitment and sitting will be on 4 addiction day hospitals, and the follow up period will be of 8 months. All subjects will have 4 sessions of group CBT (one per week) during the first month following their inclusion, and then randomly assigned (ratio 1:1) to VRET group or individual CBT group. Subjects will undergo 4 additional sessions of VRET or individual CBT (one per week) during the second month. Afterwards, all included subjects will be followed monthly for 6 months (Months 2 to 8 after inclusion). Timeline Follow-Back (TLFB) is used for the reporting of the number of alcohol standard drinks intakes, and the Transaddiction Craving Triggers Questionnaire (TCTQ) for craving assessment.