The present study characterized the discriminative-stimulus effects of triazolam and midazolam in rhesus monkeys. Six monkeys discriminated 0.1 mg/kg of triazolam from vehicle under a fixed-ratio 5 (FR 5) schedule of stimulus-shock termination (SST). Four monkeys subsequently discriminated 0.56 mg/kg of midazolam from vehicle under the same schedule of reinforcement. Benzodiazepine (BDZ) agonists midazolam and diazepam, and the barbiturate pentobarbital, substituted for triazolam, and the non-competitive N-methyl-D-aspartate (NMDA) antagonist ketamine did not. Triazolam, diazepam, lorazepam, flunitrazepam, as well as the barbiturates amobarbital and pentobarbital, substituted for midazolam, and ketamine did not. The BDZ antagonist flumazenil antagonized both the triazolam and midazolam discriminative stimuli. Bretazenil, a low-efficacy BDZ agonist, did not substitute for the midazolam discriminative stimulus in three of the monkeys and shifted the midazolam dose-effect curve to the right; in a fourth monkey, bretazenil substituted for midazolam and shifted the midazolam dose-effect curve to the left. Schild analyses with flumazenil or bretazenil, in combination with midazolam, yielded slopes that deviated significantly from unity. While clearly supporting the notion that BDZ agonists produce stimulus effects by acting at the gamma-aminobutyric acidA (GABA(A)) receptor complex, these data also suggest that the discriminative-stimulus effects of midazolam might be mediated by more than one BDZ receptor subtype.