The potencies of morphine and of the narcotic analgesic agonist-antagonists butorphanol and nalbuphine in terms of their ability to decrease enflurane MAC were studied. Following the determination of control MAC for enflurane in each dog, an intravenous bolus dose of either butorphanol tartrate, nalbuphine hydrochloride, morphine, or placebo was administered and enflurane MAC was redetermined. A higher dose of the same drug was then administered and enflurane MAC was redetermined up to a total of four doses in each animal. The successive doses for morphine and nalbuphine were 0.5, 1.5, 5.0, and 20.0 mg/kg; for butorphanol, 0.1, 0.3, 1.0, and 4.0 mg/kg; lactated Ringer's solution was used as a placebo. Both butorphanol and nalbuphine produced significant reductions of enflurane MAC (11 and 8%, respectively) at their lowest doses. No further reductions were produced by three- to forty-fold larger doses of either agonist-antagonist. Morphine produced a 17% reduction of enflurane MAC at the lowest dose with progressive decreases of enflurane MAC up to 63% at a dose of 5 mg/kg morphine. A fourfold increase in the morphine dose did not further decrease MAC. No change in enflurane MAC occurred in the animals given placebo. It was concluded that there is a "ceiling" to the potency of butorphanol and nalbuphine as anesthetic supplements. There is also a limit to the anesthetic sparing effect of morphine, but it is considerably greater than that of the agonist-antagonist narcotic analgesics.