Nonhuman and human studies comparing the abuse liability of alprazolam to other commonly used benzodiazepines are reviewed. These studies are reviewed to determine to what extent prospective, controlled, experimental studies support opinions that alprazolam's abuse liability is greater than that of other commonly used benzodiazepines. Studies comparing the self-administration of alprazolam, the discriminative stimulus effects of alprazolam, self-reported effects of alprazolam, physiological dependence on alprazolam, and adverse effects of alprazolam relative to other benzodiazepines are reviewed. Overall, the experimental literature does not support the widely held belief that alprazolam's abuse liability is greater than that of other benzodiazepines, but much more research is needed. Such research should focus explicitly on alprazolam's reinforcing effects, and the nature and severity of the discontinuation syndrome associated with its long-term use. Important issues such as selection of an appropriate comparison drug, selection of an appropriate population, dosing regimen and test doses need to be considered in future studies.