In rabbit isolated pulmonary artery previously incubated with [3H]-noradrenaline, isoprenaline (0.3 microM) had no effect on the stimulation-induced outflow of radioactivity. However, if the phosphodiesterase inhibitor ICI 63,197 (30 microM) or the alpha-adrenoceptor blocker phentolamine (1 microM) was present, then isoprenaline significantly enhanced the stimulation-induced outflow, an effect blocked by propranolol (0.1 microM). ICI 63,197 (30 microM) but not phentolamine significantly enhanced the stimulation-induced outflow of radioactivity. In mouse isolated atria previously incubated with [3H]-noradrenaline and stimulated at a frequency of 10 Hz, isoprenaline had no effect on the stimulation-induced outflow of radioactivity; this is in contrast to its release-enhancing effects at stimulation frequencies of 4 Hz and 2 Hz. The facilitation of stimulation-induced outflow by isoprenaline at 4 Hz was blocked by propranolol (0.08 microM) which, by itself, had no effect on the stimulation-induced outflow. At a stimulation frequency of 2 Hz in mouse atria the facilitatory effect of isoprenaline (0.01 microM) was significantly greater in the presence of ICI 63,197 (30 microM) which, by itself, had no effect on the stimulation-induced outflow. Similarly, the facilitatory effect of isoprenaline was significantly greater in the presence of phentolamine (1 microM) but, in this case, phentolamine significantly enhanced the stimulation-induced outflow. These results suggest that facilitatory prejunctional beta-adrenoceptors are present in both rabbit pulmonary artery and mouse atria. The effects of the phosphodiesterase inhibitor ICI 63,197 suggest that they are linked to adenylate cyclase in both tissues and we propose that the ability of phentolamine to facilitate the release and enhance the effect of isoprenaline may be due to the blockade of alpha-adrenoceptor inhibition of adenylate cyclase. This latter proposition needs further investigation.