Testicular germ cell transplantation is a novel strategy for preservation of fertility in prepubertal cancer patients, but the risk of reseeding tumor cells into cured patients presently limits clinical application of this approach. To date, no systematic evaluation of the limitations of surface marker-based decontamination of testicular samples with acute lymphoblastic leukemia has been performed. Here, surface markers for leukemic (CD4 and major histocompatibility complex class I) and germ cells (epithelia cell adhesion molecule) in testicular samples infiltrated with Roser's T-cell leukemia were identified. These markers were then used to delete leukemic cells and/or select for germ cells by flow cytometry (FACS). The resulting cell populations were analyzed by FACS, immunocytochemistry, or evaluation of leukemia transmission in syngeneic piebald variegated rats. Simple positive selection of germ cells or deletion of leukemic cells using specific surface markers was unable to effectively decontaminate testicular samples. The poor specificity of spermatogonial surface markers and aggregation of germ and leukemic cells limited the positive selection of germ cells, while immunophenotypic variation among lymphoblastic leukemia cells prevented adequate deletion of leukemic cells. Enzymatic treatment to disperse the testicular cells and feature of the intratesticular environment contributed to this immunophenotypic variation. Only germ cell selection in combination with leukemic cell deletion prevented leukemia transmission in association with intratesticular injection of the sorted cells. However, with such combined sorting, only 0.23% of the original testicular cells were recovered. With presently available techniques, flow cytometric purification of germ cells from a leukemic donor is not sufficiently effective or safe for clinical use.