SPECT imaging with [123I]iomazenil was used to measure benzodiazepine (BZ) neuroreceptor occupancy of the agonist lorazepam administered at therapeutically relevant doses in humans and supratherapeutic doses in monkeys. Lorazepam at therapeutic doses (0.03 mg/kg, i.v.) administered 90 min after the bolus injection of [123I]iomazenil had no statistically significant effect (P > 0.12) on the washout rates of regional brain activities compared to that in control subjects, although human subjects demonstrated marked sedation from the lorazepam. In baboons, the effects of higher doses of lorazepam (cumulative 0.5 mg/kg) were examined in a stepwise displacement paradigm. The in vivo potency was expressed as the ED50 (or dose required to displace 50% of receptor bound activity) and was equal to 0.34 +/- 0.01 mg/kg (mean +/- SD, n = 12). Log-logit analyses of displacement data corrected for endogenous washout showed that therapeutic doses of lorazepam were associated with < 3% BZ receptor occupancy. To examine if endogenous GABA modulates potency of the BZ agonist, the ED50 of lorazepam was compared with and without concurrent administration of tiagabine, a GABA reuptake inhibitor. These experiments were designed to measure an in vivo GABA shift of agonist potency. In vivo microdialysis demonstrated that tiagabine (up to 1 mg/kg, i.v.) increased extracellular GABA levels up to 200% of baseline, but these doses had only a minimal enhancement of lorazepam's potency to displace [123I]iomazenil. This study strongly suggests that single therapeutically relevant doses of lorazepam occupy a relatively small percentage (i.e. < 3%) of BZ receptors and that BZ binding sites have a significant (i.e. > 97%) receptor reserve.