1. The anticonflict and the rotarod impairing effects of the triazolo-benzodiazepine alprazolam (APZ) and the conventional 1,4-benzodiazepine (BDZ) diazepam (DIZ) were studied in the rat. 2. After acute administration the two drugs displayed approximately equal potency with respect to anticonflict effects in Vogel's conflict test and in Montgomery's conflict test. The results obtained in the former model indicate that also the efficacies of the two drugs are approximately equal. The shapes of the dose response curves obtained in this model differed, however, markedly. Thus, while APZ displayed a wide high efficacy dose-range (0.5-16.0 mg/kg), a maximal effect of DIZ was obtained in a narrow dose-range only (around 4.0 mg/kg). 3. Acutely, APZ was more rotarod impairing than DIZ; total impairment after 2.0 mg/kg of APZ and 3.5 mg/kg of DIZ. 4. After daily treatment with doses producing approximately equal and close to maximal anticonflict effects (APZ 1.25 mg/kg/day; DIZ 3.0 mg/kg/day) tolerance developed to the rotarod impairing (6 days of treatment) and the anticonflict (Vogel's conflict test; 9 days of treatment) effects of both drugs. It appeared, however, as if tolerance was more pronounced to the ataxic/sedative effects (approximately a four-fold rightward shift of the dose-response curve) than to the anticonflict effect (approximately a two-fold shift) of APZ, whereas tolerance developed to about the same extent to the two effects of DIZ (approximately two-fold shifts). 5. In conclusion, for APZ, but not for DIZ, after long-term administration of a dose producing marked anxiolytic-like effects, the relationship anxiolytic-like effect vs. ataxic/sedative effects was altered. These findings offer a tentative explanation to the putative advantage of APZ over conventional BDZs in the treatment of panic disorder.