The effects of the peripherally restricted opioid agonist loperamide were compared to those of morphine in the formalin test in rats. Both loperamide and morphine were efficacious in producing antihyperalgesia after both subcutaneous and intracisternal administration. The antihyperalgesic effects of peripherally administered loperamide and morphine were antagonized by both naloxone and its quaternary derivative naloxone methiodide. The effects of intracisternally administered loperamide and morphine were antagonized by naloxone SC. However, quaternary naloxone SC did not block the effects of intracisternally administered loperamide, and, quaternary naloxone blocked intracisternally morphine only at a dose approximately 10-fold higher than that required to block peripherally administered morphine. In addition, approximately 10-fold higher doses of naloxone administered SC were required to antagonize loperamide compared to doses required to antagonize morphine when the agonists were administered subcutaneously, suggesting that the effects of loperamide might be mediated by opioid receptors different from those which mediated the effects of morphine. However, neither the kappa-receptor selective antagonist nor-binaltorphimine nor the delta-receptor selective antagonist naltrindole blocked the effects of either opioid agonist. The present results are consistent with the interpretation that the antihyperalgesic effects of opioid agonists can have both a peripheral and a central component of action, and that the peripheral component of action is sufficient to produce antihyperalgesia in the formalin test after peripheral administration. The present results provide further evidence that peripherally restricted opioid agonists might provide clinically useful treatment of some pain states, in particular pain states that might involve sensitization of peripheral nociceptors.