A total of 137 microinjections of morphine sulfate in doses of 1.0-50 μg were made into various midbrain sites of 47 rats implanted with chronic indwelling cannulae. Three to six μg of morphine injected in a 0.5 μl volume produced a marked increase in hot plate reaction time when injected into a circumscribed region of the periaqueductal gray matter ventral to the cerebral aqueduct encompassing the dorsal raphe nucleus and bordering tissue. Microinjections of low doses of morphine in other midbrain loci produced less effective antinociceptive activity or no analgesia at all. Higher doses of morphine (10-50 μg) eliminated the pain response to limb pinching, but in addition caused rats to become "hyperreactive," i.e., to over respond to a variety of stimuli. Death resulted in seven of these hyperreactive rats. Although mild hyperactivity, i.e., spontaneous motor activation, was frequently observed when 3 μg of morphine was microinjected into more lateral midbrain sites, this dose elicited little or no hyperreactivity. The results of these studies are in agreement with those studies that showed identical sites of action for electrical analgesia in the rat. These results suggest morphine and electrical stimulation have similar mechanisms of action in producing analgesia when these active midbrain sites are involved.