AbstractThe translocation t(8;21)(q22;q22), which results in the fusion of the AML1 (RUNX1) and ETO (CBFA2T1) genes, is a recurrent aberration in acute myeloid leukemia (AML), preferentially correlated with FAB M2, and has the highest incidence in childhood AML. Because of the favorable prognosis, the evidence of the t(8;21) or the AML1/ETO fusion gene is mandatory in most of the therapy trials, allowing the stratification of the patients to the correct risk group in terms of treatment. Here we present six out of 59 children with AML who were positive for AML1/ETO by RT‐PCR, but showed no evidence of the classical t(8;21)(q22;q22) by conventional cytogenetics. Because of the discrepancies between molecular and cytogenetic analyses, these six patients were further investigated by fluorescence in situ hybridization analysis. Small hidden interstitial insertions resulting in an AML1/ETO rearrangement were detected in five (8.5%) of the 59 patients, whereas the sixth patient showed a cryptic three‐way translocation. The insertions could be characterized as ins(21;8) in three patients and ins(8;21) in the remaining two. Additionally, three of the patients showed secondary chromosome aberrations leading to a higher complexity of the karyotype. In conclusion, the combination of more than one standard technique in the analysis of AML1/ETO is useful to reveal the overall frequency of cryptic chromosome rearrangements and permits a better understanding of the mechanisms involved in the generation of this fusion gene. © 2003 Wiley‐Liss, Inc.