The purpose of this study was to evaluate 3 anesthetic protocols for intraduodenal drug administration by endoscopy in rhesus monkeys (Macaca mulatta). Anesthesia was induced using intramuscular ketamine and midazolam, isoflurane (inhalant gas), or intravenous propofol in male and female rhesus monkeys. A noninvasive dosing line was placed in the duodenum by use of endoscopy, and 50% dextrose (3 ml/kg) was administered. Blood pressure, heart rate, body temperature, and reflexes (corneal, palpebral, pharyngeal) and myorelaxation (mandibular reflex and reaction to limb manipulation) were evaluated every 5 min. To estimate intestinal absorption, glycemia was evaluated prior to dextrose administration and at 2, 5, 10, 15, 20, 30, 45, and 60 min after dosing. All 3 protocols resulted in successful induction of anesthesia. Recovery from isoflurane and propofol was significantly faster than from ketamine-midazolam. Duration of the recovery period after isoflurane was less variable than with propofol, but isoflurane produced greater hypothermia. Isoflurane and propofol resulted in predictable glucose absorption after intraduodenal dextrose administration, whereas ketamine-midazolam led to an inconsistent increase in glycemia.