Differences between mouse and human hearts pose a significant limitation to the value of small animal models when predicting vector behavior following recombinant adeno-associated viral (rAAV) vector-mediated cardiac gene therapy. Hence, sheep have been adopted as a preclinical animal, as they better model the anatomy and cardiac physiological processes of humans. There is, however, no comprehensive data on the shedding profile of rAAV in sheep following intracoronary delivery, so as to understand biosafety risks in future preclinical and clinical applications. In this study, sheep received intracoronary delivery of rAAV serotypes 2/6 (2 × 1012 vg), 2/8, and 2/9 (1 × 1013 vg) at doses previously administered in preclinical and clinical trials. This was followed by assessment over 96 h to examine vector shedding in urine, feces, nasal mucus, and saliva samples. Vector genomes were detected via real-time quantitative PCR in urine and feces up to 48 and 72 h post vector delivery, respectively. Of these results, functional vector particles were only detected via a highly sensitive infectious replication assay in feces samples up to 48 h following vector delivery. We conclude that rAAV-mediated gene transfer into sheep hearts results in low-grade shedding of non-functional vector particles for all excreta samples, except in the case of feces, where functional vector particles are present up to 48 h following vector delivery. These results may be used to inform containment and decontamination guidelines for large animal dealings, and to understand the biosafety risks associated with future preclinical and clinical uses of rAAV.