Intracoronary injection is a frequently used clinical protocol for stem/progenitor cell therapy to the heart. Initial donor cell retention in the heart is the key to the success of this approach; however, this process has been poorly investigated. I established an original model to quantitatively assess initial donor cell retention after intracoronary cell injection in rats using an ex-vivo heart perfusion system and investigated factors that could affect retention. The initial retention efficiency of bone marrow mononuclear cells (BMMNC) was 20% after injection into normal hearts. The majority (>90%) of BMMNC loss into the coronary effluent occurred within the first minute of injection. Increased BMMNC dose increased absolute retention with a linear dose-effect relationship, while retention efficiency was unaltered. Retention efficiency increased to 30% in hearts with ischaemia-reperfusion, while flow cytometric studies showed that surface marker expression was unchanged between the pre-injection donor cell population and the population in the coronary effluent in both normal and ischemia-reperfused hearts. This suggests that active interactions between donor cells and coronary endothelium were not critical for donor cell retention using these experimental conditions. Instead, cell size assessment revealed that larger subpopulations of BMMNC were preferentially retained compared to smaller BMMNC in both normal and ischemia-reperfused hearts. Furthermore, a larger cell type, bone-marrow derived mesenchymal stem cells (MSC; median cell size=10.1 μm), had a markedly increased retention efficiency (80%) compared to BMMNC (median cell size=7.0 μm). A greater proportion of MSC with a larger diameter were retained compared to smaller diameter MSC; this enhanced retention plateaued with MSC ≥ 9 μm. Immunohistochemical analysis using fluorescently labeled donor cells demonstrated that all BMMNC retained in normal and ischaemia-reperfusion hearts were located within the coronary vasculature, without extravasation, up to 60 minutes after injection. Heart perfusion parameters and histological features exhibited evidence of coronary embolism after intracoronary injection of MSC, but not BMMNC (up to 40x106). These data collectively suggest that passive mechanical entrapment, and not active endothelial cell-donor cell interactions, is responsible for initial donor cell retention after intracoronary injection. This has important implications for future clinical protocols of intracoronary injection of stem / progenitor cells to the heart.