Neurotensin administered intracisternally in mice increased reaction time in the hot plate test (minimum effective dose = 25 ng) and decreased writhing induced by acetic acid (minimum effective dose = 0.25 ng). At the maximum effective dose of 250 ng, the antinocisponsive action of neurotensin lasted for about 1 h. In the hot plate test, neurotensin, [Gln⁴]-neurotensin and [D-Ala²]-methionine enkephalinamide were all equiactive at the three dose levels examined, 0.00015, 0.015 and 1.5 nmoles/mouse. At antinocisponsive doses, mice treated intracisternally with neurotensin also were hypothermic and had reduced locomotor activity. Neurotensin, 2500 ng, administered intravenously did not produce any of the effects described above. D,L-5-Hydroxytryptophan, oxotremorine and chlorpromazine given systemically caused hypothermia but did not change reaction time in the hot plate procedure. Neurotensin, even at doses as high as 2500 ng given intracisternally, did not disrupt coordinated motor control or change the current-response relationship for electroshock-induced seizures. The antinocisponsive action of neurotensin in mice (hot plate test) was not antagonized by naloxone, cyproheptadine, methergoline, atropine, benzotropine, mecamylamine, chlorpheniramine, pyribenzamine, cimetidine, HEAT, propranolol, phenoxybenzamine, haloperidol or indomethacin. Neurotensin administered intracisternally was also an effective antinocisponsive agent in rats, as assessed in the hot plate procedure, but not in the tail-flick test. Naloxone did not antagonize this action of neurotensin in the rat. It is suggested that the antinocisponsive effect of neurotensin results from an interaction of the peptide with neurotensin-specific receptor sites in the central nervous system.