This study aimed to investigate immediate cell survival and distribution following different administration routes of mesenchymal stem cells (MSCs) into naturally occurring tendon injuries. Ten million MSCs, labeled with technetium-99m hexamethylpropyleneamine oxime, were implanted into 13 horses with naturally occurring tendon or ligament injuries intra-lesionally, intravenously and by regional perfusion, and traced for up to 48 h using planar gamma scintigraphy. Labeling efficiencies varied between 1.8% and 18.5% (mean 9.3%). Cells were retained in the damaged area after intra-lesional administration but only 24% of cells were still present within the tendon after 24 h. After intravenous injection, cells largely distributed to the lung fields, with no detectable cells in the tendon lesions. Significant labeling of the tendon lesions was observed in 11/12 horses following regional perfusion but at a lower level to intra-lesional injection. The highest cell numbers were retained after intra-lesional injection, although with considerable cell loss, while regional perfusion may be a viable alternative for MSC delivery. Cells did not "home" to damaged tendon in large numbers after intravenous administration. Cells were detected in the lungs most frequently after intravascular administration, although with no adverse effects. Low cell retention has important implications for designing effective clinical therapies for human clinical use. #