This study aimed to determine whether: 1) the method of cooling the hand (i.e. central [total body] vs. local surface [hand only]) influences the relationship between the adductor pollicis temperature and twitch tension; and 2) decreased evoked twitch response during hypothermia is due to reduced muscle temperature and/or the anesthetic drug used. First, the effect of local surface cooling on adductor pollicis twitch tension during isoflurane anesthesia was determined in 15 patients, while central body temperature was not allowed to decrease. Adductor pollicis temperature and twitch tension decreased in a linear manner (P< 0.05). However, the magnitude of the decreased response was only 43% of that observed during central cooling in the authors' previous study under otherwise similar experimental conditions. Second, the effect of central cooling on adductor pollicis twitch tension during nitrous oxide/fentanyl anesthesia was determined in five patients. The twitch tension did not decrease until the adductor pollicis temperature decreased below 35.2° C. Below this temperature, twitch tension decreased 16%/° C reduction in muscle temperature. These results are similar to those obtained in the authors' previous study in patients anesthetized with nitrous oxide/isoflurane anesthesia. The authors conclude that both central and local surface cooling of the adductor pollicis muscle reduces twitch tension and that the decrease in adductor pollicis twitch tension is the same during nitrous oxide/isoflurane and nitrous oxide/fentanyl anesthesia.