BACKGROUND The purpose of this study was to assess whether pharmacologic preconditioning by exogenous or endogenous adenosine prevents the deterioration of hemodynamic function and left ventricular performance during percutaneous transluminal coronary angioplasty (PTCA). Ischemic preconditioning renders the heart more resistant to subsequent ischemia. Adenosine plays a key role in its pathogenesis. Coronary angioplasty is a suitable model for the induction of myocardial ischemia. METHODS AND RESULTS We investigated 30 patients receiving PTCA of the left anterior descending coronary. Patients were randomly allocated to either dipyridamole, leading to the liberation of endogenous adenosine (0.5 mg/kg body weight, intracoronary), exogenous adenosine (20 mg intracoronary), or an equal amount of saline. Chest pain, tolerated inflation time, and ST-segment shift were registered. Left ventricular hemodynamics, isovolumetric phase indexes, indexes of volume, ejection fraction, and indexes of diastolic dysfunction were analyzed. Patients receiving endogenous or exogenous adenosine tolerated longer balloon inflation times (dipyridamole, 208 +/- 23 seconds; adenosine, 188 +/- 41 seconds; control, 153 +/- 36 seconds; P <.05). Deterioration of left ventricular ejection fraction was less severe after adenosine (72% +/- 5% before PTCA vs 64% +/- 6% during angioplasty; P =.11) and could be prevented by intracoronary dipyridamole (69% +/- 12% before PTCA vs 68% +/- 11% after PTCA; P <. 01) compared with the control group (71% +/- 7% before PTCA vs 60% +/- 7% during angioplasty). CONCLUSIONS Intracoronary application of exogenous adenosine and liberation of endogenous adenosine increase the tolerance to ischemia and prevent deterioration of left ventricular function during ischemia. These findings can be attributed to ischemic preconditioning. However, endogenous adenosine exceeds the protective effects of exogenous adenosine.