A simplified in-vivo bioassay system was used to test the carcinogenic potential of halothane in Swiss/ICR mice. Halothane was tested only at its maximum tolerated dose, and histologic examination was performed only on tumor masses and other grossly abnormal tissues found at necropsy. Two groups, each of 15 timed pregnant mice, were exposed to either halothane, 500 ppm (0.05 per cent), or compressed air for two hours on days 10--19 of pregnancy. Five days after birth the offspring were similarly exposed, three times weekly, for 78 weeks. After a ten-week, no-treatment, observation period, all remaining mice were examined by necropsy. Mice dying or killed in extremis before final sacrifice at 88 weeks of age also underwent complete gross necropsy unless extensive cannibalism or autolysis precluded examination. The incidences of malignant tumors, hepatomas or modular hyperplasias, and benign tumors in halothane-treated mice were 7, 6, and 20 per cent, respectively; there were similar incidences of these lesions in control animals. It is concluded that under the conditions of this experiment, lifetime administration of halothane at its maximum tolerated dose is not associated with an increased incidence of neoplasia in Swiss/ICR mice.