The stimulation of peripheral opioid receptors yields analgesic responses in a model of bone cancer-induced pain in mice. In order to know the type(s) of peripheral opiate receptors involved, the paw thermal withdrawal latencies were measured in C3H/HeJ mice bearing a tibial osteosarcoma, after administering selective agonists of mu-,delta-and kappa-opiate receptors. The peritumoral administration of DAGO (0.6-6 mu g) inhibited the osteosarcoma-induced hyperalgesia at doses ineffective in healthy animals, the highest one even increasing the withdrawal latencies over the control values. Naloxone-methiodide (2 mg/kg) and cyprodime (1 mg/kg), but not naltrindole (0.1 mg/kg) nor nor-binaltorphimine (10 mg/kg), antagonized DAGO-induced analgesic effects, these therefore probably being mediated through peripheral mu-opioid receptors. The peritumoral injection of DPDPE (100 mu g) induced analgesia which was inhibited by naloxone-methiodide and naltrindole but not by nor-binaltorphimine. Cyprodime partially antagonized the analgesia induced by 100 mu g of DPDPE, but did not modify the effect induced by 30 mu g of this agonist-a dose that restores the hyperalgesic latencies up to the control values. The antihyperalgesic effect induced by the peritumoral administration of U-50,488H (1 mu g) was antagonized by naloxone-methiodide and nor-binaltorphimine, but not by cyprodime nor naltrindole, thus suggesting the involvement of peripheral kappa-opioid receptors. In conclusion, the stimulation of peripheral mu-, delta- and kappa-opioid receptors is a pharmacological strategy useful for relieving this experimental type of bone cancer-induced pain, the greatest analgesic effect being achieved by stimulating peripheral mu-opioid receptors.