Purpose and Methods We reviewed the cytogenetic pattern of the malignant cells in 36 patients who were MLL abnormalities in these patients. Results Eighteen patients had 11q23 translocations or insertions detected by cytogenetic analysis (15 cases) or by FISH (3 cases); 10 patients had t(9;11), all of whom had M5a. Eight patients had other 11q23 translocations or insertions not involving chromosome 9 [t(11q23)] (four each had M4 or M5 leukemias). Eighteen cases with M4/M5 did not have 11q23 abnormalities. MLL rearrangements were found in all patients with translocations or insertions of 11q23 who were studied. Clinically, children with t(9; 11) were indistinguishable from other patients with M4-M5 leukemias. In contrast, the t(11q23) group was characterized by extreme hyperleukocytosis, CNS disease, and skin involvement. Patients with the t(9; 11) had a better outcome when compared with patients in the t(11q23) group (EFS ± SE at 3 years, 56 ± 17% versus 10 ± 10%, p = 0.04), and to all the remaining children with M4-M5 leukemias (p = 0.04). Conclusions The combination of cytogenetic, FISH, and molecular analysis provides a highly sensitive strategy for detection of 11q23/MLL gene rearrangements in childhood M4-M5 leukemias. Our more precise classification of these patients allows a more accurate correlation with outcome. The favorable prognostic significance of the t(9; 11) should be confirmed in prospective studies including a larger number of children as well as adults.