I nhibitory M2 muscarinic receptors on parasympathetic nerve endings in the lungs decrease release of acetylcholine, inhibiting vagally induced bronchoconstriction. Neuronal M2 receptor function can be studied using selective agonists and antagonists such as pilocarpine and gallamine. In pathogen-free guinea pigs indomethacin (1 mg/kg) did not alter the effect of either gallamine or pilocarpine, thus in pathogen free animals neuronal M2 muscarinic receptors function independently of cyclooxygenase products. However, in guinea pigs infected with virus, (which causes temporary loss of M2 receptor function), and then allowed to recover for 8 wk (to allow recovery of M2 receptors), indomethacin prevented both gallamine's potentiation and pilocarpine's inhibition of vagally induced bronchoconstriction. This new effect of indomethacin was not blocked by the addition of a 5-lipoxygenase inhibitor, AA861. However, the selective COX II inhibitor, L-745,337, had the same effect as indomethacin. Since exposure to ozone also caused neuronal M2 receptors to become dependent upon cyclooxygenase the effects of viral infection are likely to be due to inflammation. Thus, despite apparent recovery of normal M2 receptor function after viral infection or ozone, linkage of these receptors is chronically altered such that they become largely dependent on the activity of COX II.