Although preclinical studies suggest that GHB has low likelihood for abuse, case reports indicate that GHB is abused. This study evaluated the relative abuse liability of GHB in 14 volunteers with histories of drug abuse. Psychomotor, subjective, and cognitive effects of a broad range of GHB doses (2-18 g/70 kg), up to a dose that produced severe behavioral impairment in each participant, were compared to placebo and two abused sedative/hypnotic drugs, triazolam (0.5 and 1 mg/70 kg) and pentobarbital (200 and 400 mg/70 kg), under double-blind, double-dummy conditions at a residential research facility. In general, GHB produced effects similar to triazolam and pentobarbital, although GHB was not identified as a benzodiazepine or barbiturate by participants that correctly identified triazolam and pentobarbital as such. On most measures of likelihood of abuse (eg ratings of liking, reinforcing effects), effects of pentobarbital were significantly greater than those of triazolam, with GHB being intermediate. GHB produced significantly greater negative subjective effects, including nausea, than the other drugs. Memory impairment after GHB was less than that after triazolam and pentobarbital. Within participants, the dose-effect function for sedation was steeper for GHB than for triazolam and pentobarbital. Also, at higher doses, GHB was associated with greater sedation and more variability across participants in sedation. Taken together, these data suggest that the profile of effects of GHB only partially overlaps with that of triazolam and pentobarbital. Although the likelihood for GHB to be abused is intermediate to triazolam and pentobarbital, the possibility of accidental overdose (ie greater sedation than intended) with GHB appears to be greater.