AIMS Contradictory outcomes from recent clinical trials investigating the transplantation of autologous bone marrow mononuclear cell (BM-MNC) fraction containing stem/progenitor cells to damaged myocardium, following acute myocardial infarction, may be, in part, due to the different cell isolation protocols used. We compared total BM-MNC numbers and its cellular subsets obtained following isolation using Ficoll-Paque and Lymphoprep - two different density gradient media used in the clinical trials. MATERIALS & METHODS Bone marrow samples were taken from patients entered into the REGENERATE-IHD clinical trial after 5 days of subcutaneous granulocyte colony-stimulating factor injections. Each sample was divided equally for BM-MNC isolation using Ficoll-Paque and Lymphoprep, keeping all other procedural steps constant. Isolated fractions were characterized for hematopoietic stem cells, endothelial progenitor cells, T lymphocytes, B lymphocytes and NK cells using cell surface markers CD34(+), CD133(+)VEGFR2(+), CD45(+)CD3(+), CD45(+)CD19(+) and CD45(+)CD16(+)CD56(+), respectively. There were no significant differences in the absolute numbers and percentage cell recovery of various mononuclear cell types recovered following separation using either density gradient media. Cell viability and the proportion of various cell phenotypes investigated were similar between the two media. They were also equally efficient in excluding unwanted red blood cells, granulocytes and platelets from the final cell products. CONCLUSION We demonstrated that the composition and quantity of cell types found within therapeutic BM-MNC preparations for use in clinical trials of cardiac stem cell transplantation are not influenced by the type of density gradient media used when comparing Ficoll-Paque and Lymphoprep.