Baclofen has been suggested as a potential pharmacotherapy for alcohol use disorder, but the clinical data are conflicting. Here we investigated the biobehavioral effects of baclofen in a sample of anxious alcohol-dependent individuals. This was a randomized, double-blind, placebo-controlled, human laboratory study in non-treatment seeking alcohol-dependent individuals with high trait anxiety (N=34). Participants received baclofen (30 mg per day) or placebo for at least 8 days, then performed an experimental session consisting of alcohol cue-reactivity followed by alcohol administration procedure (alcohol priming, then alcohol self-administration). Total amount of alcohol self-administered was the primary outcome; alcohol craving, subjective/physiological responses and mood/anxiety symptoms were also evaluated. There was no significant medication effect on the total amount of alcohol consumed during the alcohol self-administration (P=0.76). Baclofen blunted the positive association between maximum breath alcohol concentration during priming and the amount of alcohol consumption (significant interaction, P=0.03). Ratings of feeling intoxicated were significantly higher in the baclofen group after consuming the priming drink (P=0.006). During the self-administration session, baclofen significantly increased ratings of feeling high (P=0.01) and intoxicated (P=0.01). A significant reduction in heart rate (P<0.001) and a trend-level increase in diastolic blood pressure (P=0.06) were also detected in the baclofen group during the alcohol laboratory session. In conclusion, baclofen was shown to affect subjective and physiological responses to alcohol drinking in anxious alcohol-dependent individuals. These results do not support an anti-craving or anti-reinforcing effect of baclofen, but rather suggest that baclofen may act as a substitution medication for alcohol use disorder.