The effects of the major morphine metabolites, morphine-3-glucuronide and morphine-6-glucuronide, on nociception were assessed by the tail-flick, hot-plate and writhing tests in the rat. Morphine-3-glucuronide (M3G) 1.1 x 10(-9) mol (0.5 micrograms) or saline was injected intracerebroventricularly (i.c.v.) or intrathecally (i.t.) followed by a second injection of 2.0 x 10(-10) mol (0.1 microgram) or 2.0 x 10(-11) mol (0.01 microgram) morphine-6-glucuronide (M6G) 10 min later. Administration of M3G (i.c.v.) significantly attenuated the antinociceptive effects of M6G in the hot-plate test. After i.t. administration, the antinociceptive effect of M6G in all three tests was significantly reduced in the M3G pretreated group compared to the group receiving saline. The ventilatory effects of 4.0 x 10(-9)-1.0 x 10(-8) mol (2-5 micrograms) M6G and 1.7-2.2 x 10(-8) mol (8-10 micrograms) M3G given i.c.v. were studied by a whole-body plethysmographic technique in halothane anaesthetized rats. Separate groups of rats received M3G followed by M6G injection or vice versa. In animals receiving M3G there was a prevention or attenuation of the M6G induced depression of respiratory frequency, tidal volume and minute ventilation compared to control groups receiving M6G in combination with saline. These results show that M3G may functionally antagonize the central antinociceptive effects as well as the ventilatory depression induced by M6G. Interestingly, M3G was more potent in antagonizing the M6G-induced analgesia after i.t. administration than that after i.c.v. administration, which may suggest that the spinal cord is more sensitive to the non-opioid excitatory effects of M3G than supraspinal structures.