1 A modified abdominal constriction test, whereby the drugs used are injected intraperitoneally when the writhing response is maximal, has been used to study the antinociceptive activity of various sympathomimetic drugs. Of those tested, clonidine was the most potent, with an ID(50) value in the nanomolar range. (-)-Isoprenaline, (-)-adrenaline and (-)-noradrenaline were only a little less potent. Phenylephrine, the least potent, had only about one-sixtieth of the activity of clonidine.2 The antinociceptive action appears to occur within the peritoneum, since it was apparent almost immediately after the drugs were injected and was produced by doses far smaller than were effective by the subcutaneous route.3 alpha-Adrenoceptors appear to be involved in the reaction, since noradrenaline showed stereospecificity, and the alpha-adrenoceptor antagonists phentolamine and piperoxan both shifted the dose-response curves of the alpha-adrenoceptor agonist drugs to the right, usually parallel to the control curves.4 The high antinociceptive potency of clonidine and oxymetazoline, indicate the importance of alpha(2)-adrenoceptors and this was supported by the finding that piperoxan was a more effective antagonist than phentolamine. The moderate potency of phenylephrine suggests that alpha(1)-adrenoceptors may also be involved, although the selective alpha(1)-antagonist, prazosin, did not antagonize noradrenaline and had antinociceptive activity of its own.5 beta-Adrenoceptors also appear to be involved in the antinociceptive response, since propranalol antagonized the effect of isoprenaline, but not that of clonidine.6 Piperoxan was a very effective antagonist of morphine, while phentolamine had a weaker action. Naloxone had little action against the alpha-adrenoceptor agonists.7 Mice pretreated with clonidine or oxymetazoline but not noradrenaline showed a very great cross-tolerance to morphine. Morphine pretreatment caused marked desensitization of itself, but little cross-tolerance to clonidine or oxymetazoline.8 It is suggested that sensory nerves in the mouse peritoneum have alpha(2)- and beta-adrenoceptors on their terminals, and possibly alpha(1)-receptors also. It is possible that when activated by the appropriate agonists they depress the generation of pain impulses. There is an interaction between the alpha-adrenoceptors and opioid receptors in the mouse peritoneum.