Alprazolam was studied for its effects on schedule-controlled behavior of squirrel monkeys. Monkeys responded by pressing a lever under fixed-ratio or two separate fixed-interval (FI) schedules. Responding was maintained by the presentation of food under all three schedules and was suppressed under one of the FI schedules by electric shock. Unlike lorazepam, alprazolam did not increase rates of either suppressed or nonsuppressed responding under the two FI schedules after acute administration. When given for 3 (at 0.03 mg/kg) or 11 (at 0.1 mg/kg) consecutive days immediately before test sessions, alprazolam also did not increase suppressed responding. Under all three schedules, alprazolam (0.03-3.0 mg/kg) produced dose-dependent decreases in response rate. The rate-decreasing effects of alprazolam were attenuated by drugs known to have benzodiazepine (BZ) antagonist action: Ro 15-1788 (0.01 and 0.3 mg/kg), Ro 15-4513 (0.01 and 0.1 mg/kg) and 3-(methoxycarbonyl)amino-beta-carboline (10.0 and 30.0 mg/kg). These latter drugs alone either had no effect on or decreased fixed-ratio responding. Alprazolam also was tested in combination with lorazepam under the two FI schedules. In general, low doses of lorazepam, equal to or less than those having little effect on response rate when given alone, increased suppressed and nonsuppressed responding when combined with low to moderate doses of alprazolam. The results show that although the rate-decreasing effects of alprazolam probably are due to actions at BZ recognition sites, alprazolam alone does not have typical BZ-like agonist effects on FI responding; it did, however, potentiate the behavioral effects of lorazepam.