1. The effects of isoenzyme-selective inhibitors of phosphodiesterases PDE3 and PDE4 on cyclic AMP concentration, two indices of phosphoinositide hydrolysis, and contractile responses to spasmogens have been investigated in bovine tracheal smooth muscle (BTSM). 2. Neither the PDE3-selective inhibitor ORG 9935, nor the PDE4-selective inhibitor rolipram increased cyclic AMP levels in BTSM. However, rolipram addition in the presence of PDE3 inhibition (ORG 9935; 1 microM) concentration-dependently (-log EC50 (M), 6.55+/-0.15; n = 3) increased cyclic AMP levels to about 70% of the maximal response to the beta-adrenoceptor agonist isoprenaline. 3. Rolipram per se inhibited histamine-stimulated [3H]-inositol (poly)phosphate ([3H]-InsP(X)) accumulation by > 80% (-log EC50 (M), 6.92+/-0.11; n = 3). Although ORG 9935 (1 microM) had little effect on histamine-stimulated [3H]-InsP(X) accumulation alone it greatly facilitated the inhibitory action of rolipram (-log EC50 (M), 8.82+/-0.39; n = 3). The effects of PDE3 and/or PDE4 inhibition on [3H]-InsP(X) accumulation stimulated by muscarinic acetylcholine (mACh) receptor activation were less marked. However, combined PDE3/4 inhibition significantly decreased this response at a submaximal concentration of mACh receptor agonist (carbachol; 1 microM). 4. The greater-than-additive effect of combined PDE3/4 inhibition was also observed at the level of contractile responses to histamine and carbachol. In experiments designed to investigate the effects of PDE3 and/or 4 inhibitors on the carbachol-mediated phasic contraction, additions of rolipram (10 microM) or ORG 9935 (1 microM) were without effect, whereas added together the inhibitors caused a significant (P < 0.01) 40% reduction in the peak phasic contractile response. 5. The effect on contraction correlated with a substantial inhibitory effect of PDE3/4 inhibition on the initial increase in inositol 1,4,5-trisphosphate (InsP3) accumulation stimulated by spasmogen. Thus, in the presence of ORG 9935 (1 microM) rolipram concentration-dependently inhibited carbachol-stimulated InsP3 accumulation by > or = 50% (-log EC50 (M), 6.77+/-0.21; n = 4). 6. Carbachol (100 microM) addition caused a rapid decrease (by 67% at 10 s) in BTSM cyclic AMP level in the presence of PDE3/4 inhibition. However, omission of Ca2+ from the incubation medium prevented the carbachol-evoked decrease in cyclic AMP and this coincided with a greater inhibition (> or = 80%) of the carbachol-stimulated InsP3 response. 7. These data indicate that combined PDE3 and PDE4 inhibition has greater-than-additive effects on second messenger and functional responses to spasmogens in BTSM. Furthermore, the ability of PDE3/4 inhibition significantly to attenuate mACh receptor-mediated contractile responses, may be, at least in part, attributed to an effect exerted at the level of InsP3 generation.