Gene transfer with replication-deficient recombinant adenovirus (Ad) vectors may provide a novel approach to the treatment of some cardiac disorders. The relative efficiency of intramyocardial vs intracoronary Ad vector injection in transducing myocardial cells remains to be determined. Further, Ad vectors are associated with localized inflammation, and this could be associated with clinically significant side-effects. Female minipigs underwent open chest surgery and the Ad vector AdCMV. NLS beta-gal was injected into the circumflex coronary artery (IC; 2 x 10 (10) pfu; n= 5) or the posterobasal wall of the left ventricle (im; 5 x 10 (9) pfu, n= 4; 2 x 10 (10) pfu, n= 18). The minipigs were killed after 2-31 days and the hearts examined for evidence of beta-galactosidase activity. Minipigs underwent epicardial echocardiography immediately before, within 15 min following the im injection of AdCMV. NLS beta-gal and again at the time of death. Blood samples for white blood cell count, alkaline phosphatase, total bilirubin, blood urea nitrogen, creatinine and electrolytes were obtained before im and ic injection of the Ad vector and before death. Intramuscular injection of the Ad vector was more efficient than ic infusion in infecting cells in a localized area of the heart. Myocardial beta-gal activity peaked at 3-6 days after im injection and returned to its control value within 1 month. Although inflammatory cells were present at the injection site, echocardiograms did not show any evidence of either segmental or global left ventricular dysfunction. No minipigs died and all blood tests remained within normal limits following either im or ic exposure to the Ad vector. In summary, direct im administration of replication-deficient, recombinant Ad vectors provides a safe and effective approach for short-term gene transfer into the heart of large mammals.