The chronic i.c.v. infusion of morphine has been reported for rats but not for mice. In the current report, the antinociceptive tolerance to both i.c.v. morphine infusion and s.c. implantation of morphine pellets in mice was compared. Physical dependence after i.c.v. morphine infusion was also evaluated. Osmotic minipumps were filled with morphine (50 mM), connected to i.c.v. cannulae, and implanted s.c. to deliver 50 nmol/h for 3 days (i.e., 3.6 mu mol total). Robust jumping precipitated by naloxone (I mg/kg, s.c.) indicated the development of physical dependence. Tolerance to i.c.v., i.t., and i.v. morphine (6.3-, 2.0-, and 4.4-fold, respectively) was observed using the tail flick test. Mice implanted with pellets containing 75 mg morphine for 3 days (i.e., similar to 260 mu mol total) were also tolerant to morphine (6.5-, 7.5- and 18-fold, respectively). Thus, the tolerance developed using the two methods was not identical. These results allow comparison of morphine tested by 3 different routes (i.c.v., i.t., and i.v.) after chronic morphine treatment by two routes (i.c.v. and s.c.) in a single study. (c) 2005 Elsevier B.V. All rights reserved.