Three major analgesics (morphine, pentazocine, methotrimeprazine) were tested on intact and decerebrate mice for their ability to prolong the reaction time of motor nociceptive responses in two established analgesimetric tests: the thermal receptacle and the hot-plate method. In intact mice only morphine showed a marked dose-dependent anti-nociceptive activity by both methods. Methotrimeprazine ranked higher than morphine, as measured by its ED50, in the hot-plate method, but it showed an early ceiling effect in the thermal receptacle technique. Pentazocine showed a very early ceiling effect in both tests. Transections tangent to the superior colliculi did not significantly modify the mean reaction times recorded in either of the two methods, though they caused a significant change in the pattern of response of mice tested by the hot-plate technique. Drug parameters, as ED50's and slope functions, remained unchanged following decerebration, except for a significant drop in the potency of morphine as determined by the thermal receptacle method. The relevance of these findings to mechanisms of analgesic action and the reliability of analgesic screening are discussed.