In order to evaluate and to compare the time course, dose response, and the degree of tolerance development to butorphanol and morphine, rats were continuously intracerebroventricularly (ICV) infused with saline vehicle (1 microliter/h), butorphanol (6.5, 13, 26, and 52 nmol/microliters/h), or morphine (1.6, 6.5, and 26 nmol/microliters/h) through osmotic minipumps for 1 to 3 days. The tail-flick responses were determined pre-, during, and postinfusion. Tolerance to morphine developed faster than that to butorphanol. The antinociceptive response to the ICV challenge dose (6 h after the termination of drug infusion) of butorphanol or morphine was decreased significantly and there was a negative correlation between the dose of the drug infused and the observed antinociceptive response. In terms of butorphanol and morphine tolerance, a parallel rightward shift in the dose-response curve was produced with the degree of shift proportional to the log of the infusion dose. In tail-flick tests, the shifts of the dose-response curves for butorphanol and morphine in tolerant animals were 11.8- and 46.3-fold, respectively. However, in the acetic acid writhing test, the shifts of the dose-response curves for butorphanol and morphine in tolerant animals were 11.3- and 11.7-fold, respectively. These results suggest that there is a greater degree of tolerance to morphine than there is to butorphanol, but the degree of butorphanol tolerance is still substantial. In addition, two pain assays (tail flick vs. writhing) yielded different estimations of tolerance in a comparison of morphine and butorphanol.