Mechanisms of adenosine (ADO) protection of reperfused myocardium are not fully understood. We tested the hypothesis that ADO (0.1 mM) alleviates ventricular stunning by ADO A(1)-receptor stimulation combined with purine metabolic enhancements. Langendorff guinea pig hearts were stunned at constant left ventricular end-diastolic pressure by low-flow ischemia. Myocardial phosphate metabolites were measured by (31)P-NMR, with phosphorylation potential {[ATP]/([ADP].[P(i)]), where brackets indicate concentration} estimated from creatine kinase equilibrium. Creatine and IMP, glycolytic intermediates, were measured enzymatically and glycolytic flux and extracellular spaces were measured by radiotracers. All treatment interventions started after a 10-min normoxic stabilization period. At 30 min reperfusion, ventricular contractility (dP/dt, left ventricular pressure) was reduced 17-26%, ventricular power (rate-pressure product) by 37%, and [ATP]/([ADP].[P(i)]) by 53%. The selective A(1) agonist 2-chloro-N(6)-cyclo-pentyladenosine marginally preserved [ATP]/([ADP].[P(i)]) and ventricular contractility but not rate-pressure product. Purine salvage precursor inosine (0.1 mM) substantially raised [ATP]/([ADP].[P(i)]) but weakly affected contractility. The ATP-sensitive potassium channel blocker glibenclamide (50 microM) abolished ADO protection of [ATP]/([ADP].[P(i)]) and contractility. ADO raised myocardial IMP and glucose-6-phosphate, demonstrating increased purine salvage and pentose phosphate pathway flux potential. Coronary hyperemia alone (papaverine) was not cardioprotective. We found that ADO protected energy metabolism and contractility in stunned myocardium more effectively than both the A(1)-receptor agonist 2-chloro-N(6)-cyclo-pentyladenosine and the purine salvage precursor inosine. Because ADO failed to stimulate glycolytic flux, the enhancement of reperfusion, [ATP]/([ADP].[P(i)]), indicates protection of mitochondrial function. Reduced ventricular dysfunction at enhanced [ATP]/([ADP].[P(i)]) argues against opening of mitochondrial ATP-sensitive potassium channel. The results establish a multifactorial mechanism of ADO antistunning, which appears to combine ADO A(1)-receptor signaling with metabolic adenylate and antioxidant enhancements.