The antinociceptive effects of morphine-6-glucuronide (M6G) were examined in two animal models of pain, the tail immersion test (reflex withdrawal to noxious heat) and the formalin test (behavioral response to minor tissue injury). In the tail immersion test, M6G produced an increase in withdrawal latency that rose rapidly between 0.01 and 0.025 ug ICV or 1 and 2 mg/kg SC. A further increase occurred at doses greater than 0.2 ug ICV or 4 mg/kg SC and was associated with marked catalepsy and cyanosis. Naloxone, 0.1 mg/kg SC, shifted the lower component of the dose-effect relation by a factor of 24. In the formalin test, 0.01 ug M6G ICV produced hyperalgesia, while between 0.05 and 0.2 ug ICV, antinociception increased rapidly without toxicity. The dose effect relations for hyperalgesia and antinociception were shifted to the right by factors of 20- and 3-fold, respectively. By comparison, ICV morphine was 60 (formalin test) to 145-200 (tail immersion test) times less potent than M6G. At sub-nanomolar concentrations, M6G enhanced the binding of [3H]-etorphine, [3H]-dihydromorphine and [3H]-naloxone to rat brain membrane receptors by 20-40%. At higher concentrations, M6G displaced each ligand from binding sites, with Ki values of about 30 nM, as compared to morphine Ki values of about 3 nM. The data indicate that the in vivo and in vitro effects of M6G are complex and that M6G may play an important role in analgesia in experimental animals, and by implication, in man.