A non-contracting scar following myocardial infarction can adversely affect ventricular topography and hemodynamic function. Gene transfer has the potential to prevent or alter such pathophysiological processes. Normal myocardium is a proven target for delivery of DNA or viral vectors but the potential for gene therapy in ischemic myocardium has not been evaluated. In an initial series of experiments, we determined whether the direct injection of reporter genes into hearts subjected to coronary artery occlusion followed by reperfusion could result in gene expression comparable to the levels observed in non-occluded normal hearts. Anesthetized rats were subjected to 15 min or 60 min of proximal coronary occlusion or sham operation. Luciferase gene under the control of the Rous sarcoma virus promoter was injected directly into the anterior left wall. At 1 week, high expression of luciferase was observed in both the ischemic/reperfused and non-ischemic tissue. Thus DNA transfer by direct injection is possible after ischemic injury and uptake and expression are not impaired. In a second series of experiments, myocardial infarcts in dogs were injected with a beta-galactosidase expressing retroviral vector. LNPOZ. Six to 11 days later frozen sections revealed macroscopically visible expression of beta-galactosidase activity. Not only can foreign genes be taken up by direct injection of DNA or retroviruses into ischemic/reperfused myocardium but they can be transcribed and the protein synthetic machinery of the injured cells can produce recombinant polypeptides that retain enzymatic activity. These results open the way for the investigation of gene therapy in models of ischemia.