In the present study the hypothesis was tested that local noradrenaline release contributes to adenosine formation in myocardial ischemia. Therefore, in ischemic non-working rat hearts either adrenergic receptors or ischemia-evoked noradrenaline release were blocked. Noradrenaline and adenosine were determined in the effluent using HPLC-methods. Following 20 min of stop of perfusion flow both the beta-adrenergic receptor antagonist bisoprolol (91.6 +/- 10.5 nmol/g) and the inhibitor of ischemia-induced noradrenaline release desipramine (108.5 +/- 12.5 nmol/g) caused a suppression of adenosine release (control: 140.9 +/- 7.3 nmol/g). To examine the time-course of the release, further experiments were performed at constant perfusion flow with energy metabolism blocked by cyanide together with removal of glucose from the perfusion buffer. This condition resulted in a nearly simultaneous release of adenosine and noradrenaline from the hearts. The beta-adrenoceptor blocking agents atenolol and bisoprolol postponed the release of adenosine, whereas the alpha-antagonists prazosin and yohimbine had no effect on adenosine release induced by cyanide. None of the adrenergic receptor blockers affected the release of noradrenaline. The inhibitors of the neuronal noradrenaline carrier (uptake1) desipramine, oxaprotiline, and cocaine suppressed the release of noradrenaline during cyanide administration, indicating a carrier-mediated efflux of noradrenaline. Reduction of extracellular noradrenaline by these agents coincided with a delay of adenosine release (cumulative release within 20 min--control: 251.2 +/- 13.9, desipramine: 172.1 +/- 15.3, oxaprotiline 36.5 +/- 5.8, cocaine: 111.8 +/- 23.6 nmol/g). Desipramine and cocaine were also used during administration of exogenous noradrenaline in normoxic hearts, to confirm specificity of their action.(ABSTRACT TRUNCATED AT 250 WORDS)