Various arguments suggest that CD8+ T lymphocytes play a major role in the control of cytomegalovirus (CMV) infection. The detection of CMV-specific CD8+ T cells may therefore provide additional information about CMV virus detection to predict the risk of development of CMV disease, especially in immunodepressed transplant recipients. We compared and tested various experimental conditions to optimize an enzyme-linked immunospot assay (Elispot) assay for the detection of CMV-specific CD8+ T lymphocytes. The indirect Elispot assay with one six-day in vitro sensitization step was found to be the most sensitive method to detect CMV-specific CD8+ T cells compared to direct Elispot with unfractionated peripheral blood mononuclear cells or purified CD8+ T cells. We showed that low doses of interleukin-2 during the in vitro culture enhanced the sensitivity of this test, and tetramer staining was performed to verify the high efficiency of this in vitro stimulation step. We directly loaded the specific CMV peptide during the Elispot assay and demonstrated that the use of T2 cells did not improve its sensitivity. Elispot for the detection of interferon-gamma appears to be more sensitive and reliable than measurement of tumor necrosis factor alpha or granzyme B. This technique was successfully applied to detect CMV-specific CD8+ T cells in human leukocyte antigen A2 (HLA-A2) and HLA-B7 healthy patients and in one lymphopenic post-transplant patient with positive CMV serology. This highly sensitive test may be a useful tool to assess T-cell immunity directed against CMV in immunodepressed patients.