1. It is now accepted that uncontrolled proliferation of human airway smooth muscle (HASM) cells contributes, in many cases, to the chronic stages of asthma. However, the physiological and pathophysiological processes regulating cell growth and division in the airway are not clear. We have recently shown that the immediate early gene, cyclo-oxygenase-2, is induced by cytokines in HASM cells. Since cyclo-oxygenase metabolites, such as prostaglandin (PG) E2 have been shown to modulate HASM cell growth, we have investigated any autocrine action of endogenously released cyclo-oxygenase-1/2 products on the proliferative responses in these cells. 2. HASM cells were cultured from healthy tissue obtained at lung or heart/lung transplantation. HASM cell proliferation was measured by [3H]-methyl thymidine uptake by cells and by cell counts. Cyclo-oxygenase-2 expression was measured by Western blot analysis and activity measured by the release of PGE2, by radioimmunoasay. 3. HASM cells proliferated in response to foetal calf serum, a response that was greatly inhibited when cyclo-oxygenase-2 was induced with either interleukin-1beta plus tumour necrosis factor-alpha or interleukin-1beta, tumour necrosis factor alpha plus interferon gamma (each at 10 ng ml(-1)). The inhibitory effect of cytokines on HASM cell proliferation was reversed in a concentration dependent manner by either the mixed cyclo-oxygenase-1/-2 inhibitor, indomethacin or the selective cyclo-oxygenase-2 inhibitor, L-745,337 (each at 10 microM). 4. PGE2 or the stable analogue of prostacyclin, cicaprost concentration-dependently (0.1 pmol to 1 microM) inhibited serum induced proliferation of HASM cells. By contrast, the TP receptor agonist, U46619 stimulated proliferation of HASM cells when cells were cultured without but not with serum. Other cyclo-oxygenase products, PGD2, PGF2alpha had no effect on cellular proliferation at concentrations up to 1 microM. 5. These observations illustrate a profound inhibitory effect of cyclo-oxygenase-2 induction on HASM cell proliferation, possibly via IP or EP receptor activation. Cyclo-oxygenase-2 induction has, thus far, been associated with the pro-inflammatory responses of plasma exudation and oedema formation and is assumed to be an enzyme worthy of selective inhibition in many disease states. However, our observations suggest that cyclo-oxygenase-2 can have an anti-inflammatory, anti-proliferative function in the airways. These observations may have importance in the use and development of therapies for airway disease such as asthma.