For several decades, studies conducted to evaluate the efficacy of RS(±)‐Baclofen in the treatment of alcohol dependence yielded contrasting results. Human and animal studies recently questioned the use of the racemic drug in patients since a potential important role of the different enantiomers has been revealed with an efficacy thought to reside with the active R(+)‐enantiomer. Here we conducted experiments in the postdependent rat model of alcohol dependence to compare the efficacy of R(+)‐Baclofen or S(−)‐Baclofen to that of RS(±)‐Baclofen on ethanol intake, seeking, and relapse. R(+)‐Baclofen was more effective than RS(±)‐Baclofen in reducing ethanol intake and seeking during acute withdrawal and during relapse after abstinence. We also used an original population approach in order to identify drug responders. We found a significant proportion of responders to S(−)‐Baclofen and RS(±)‐Baclofen, displaying an increase in ethanol intake, and this increasing effect on alcohol intake was not seen in the R(+)‐Baclofen group. At an intermediate dose of R(+)‐Baclofen, devoid of any motor side effects, we identified a very large proportion of responders (75%) with a large decrease in ethanol intake (90% decrease). Finally, the response to RS(±)‐Baclofen on ethanol intake was correlated to plasma level of Baclofen. R(+)‐Baclofen and RS(±)‐Baclofen were effective in reducing sucrose intake. Our study has important clinical implication since it suggests that the wide variability in the therapeutic responses of patients to RS(±)‐Baclofen may come from the sensitivity to the R(+)‐Baclofen but also to the one of the S(−)‐Baclofen that can promote an increase in ethanol intake.