This study determined the anesthetic efficacy of midazolam (MID) in terms of its ability to reduce enflurane MAC (EMAC). Control EMAC was determined by the tail-clamp method in 15 mongrel dogs. Each animal then received at least three incremental infusion rates of MID from among the following: 0.48, 2.4, 9,6, 19.2, 28.8, 48, or 151.2 μg · kg−1 · min−1. MAC was determined during each infusion rate following a 1-h observation period, during which time MID concentration in plasma ([MID]) stabilized. [MID] was measured every 15 min beginning 45 min from the start of each new infusion rate. There was a linear relationship between MID infusion rates and the resulting [MID] (r = 0.995). In the range of [MID] from 14 to 14,118 ng/ml, there was a linear relationship between the log [MID] and the percent EMAC reduction. The slope of the line was very shallow, and the [MID] required to reduce EMAC by more than 50% exceeded the [MID] likely to be employed clinically in humans (750 ng/ml). Also, the 73 ± 4% (mean ± SEM) EMAC reduction produced by [MID] = 9,763 ± 1213 ng/ml was not significantly greater than the 60 ± 3% EMAC reduction achieved by [MID] = 1,464 ± 293 ng/ml, a finding which suggests a ceiling effect to the anesthetic efficacy of midazolam. The authors conclude that, within the dose range of MID likely to be employed in humans, MID produced a concentration-dependent reduction of enflurane MAC in the dog. In doses above those likely to be employed clinically, a ceiling effect to the anesthetic efficacy of MID may become evident.