One important question in stem cell biology of childhood acute lymphoblastic leukemia (ALL) is whether immature CD34+CD19- cells are part of the leukemic cell clone. CD34+CD19- cells from the bone marrow of 9 children with TEL/AML1-positive ALL were purified by flow sorting and subjected to reverse transcriptase-polymerase chain reaction (RT-PCR), fluorescence in situ hybridization, and methylcellulose cultures. In 3 of 8 patients analyzed by RT-PCR, no TEL/AML1-positive cells could be detected in the CD34+CD19- cell fraction. Altogether, the percentage of TEL/AML1-positive cells was low: 1.6% (n = 8; SD 2.2%) by nested real-time RT-PCR and 2.5% (n = 5; SD 2.6%) by fluorescence in situ hybridization. This correlated with the percentage of contaminating CD19+ leukemic cells in the CD34+CD19- cell fraction in 6 control sorts (mean 4.6%, SD 3.6%), indicating that the low levels of leukemic cells detected in the CD34+CD19- cell fraction could be attributed to sorter errors. Methylcellulose cultures in 3 patients provided further evidence that CD34+CD19- cells represent a candidate normal cell population. The clonogenicity of the CD34+CD19- cell fraction was similar to normal progenitors, including growth of primitive granulocyte, erythroid, macrophage, megakaryocyte colony-forming units. Each of 92 colonies from cultures with CD34+CD19- cells tested negative for TEL/AML1. In conclusion, our data support the hypothesis that the leukemia in TEL/AML1-positive childhood ALL originates in a CD19+ lymphoid progenitor. This has many therapeutic implications, eg, for purging of autologous stem cell products, flow cytometric monitoring of minimal residual disease, and targeting immunotherapy against the leukemic cell clone.