In drug discrimination (DD) procedures, behavior is differentially reinforced depending on the presence or absence of specific drug stimuli. The DD paradigm has been widely adopted by behavioral pharmacologists because of its specificity of stimulus control, concordance with drug action at cellular levels and its use as a preclinical model of subject-rated effects in humans. With the successful extension of DD to humans, a comparison of human and nonhuman DD will help place each in the context of the other. Twenty-eight studies of DD in humans are reviewed, including studies of amphetamine, opioid, benzodiazepine, caffeine, nicotine, marijuana and ethanol discriminative stimuli. Comparison of procedures between studies in humans and nonhumans reveals a common tradition, except the use of instructions appears to facilitate greatly DD acquisition in humans. Findings were qualitatively similar between humans and nonhumans. Potency relationships were quantitatively similar between humans and most, but not all, other species. Areas of human DD needing additional empirical evaluation include the influence of instructions, the effects of training dose and the effects of antagonists. Additionally, antihistamines, barbiturates, nicotine and marijuana are under-represented in human DD.