Chromosomal insertion-derived BCR-ABL1 fusion is rare and mostly cryptic in chronic myeloid leukemia (CML). Most of these cases present a normal karyotype, and their risk and/or prognostic category are uncertain. We searched our database and identified 41 CML patients (20 M/21 F, median age: 47 years, range 12-78 years) with insertion-derived BCR-ABL1 confirmed by various FISH techniques: 31 in chronic phase, 1 in accelerated phase, and 9 in blast phase at time of diagnosis. Conventional cytogenetics analysis showed a normal karyotype (n = 19); abnormal karyotype with morphologically normal chromosomes 9 and 22 (n = 5); apparent ins(9;22) (n = 2) and abnormal karyotype with apparent abnormal chromosomes 9, der(9) and/or 22, der(22) (n = 15). The locations of insertion-derived BCR-ABL1 were identified on chromosome 22 (68.3%), 9 (29.3%), and 19 (2.4%). Complex chromosomal abnormalities were often overlooked by conventional cytogenetics but identified by FISH tests in many cases. After a median follow-up of 58 months (range 1-242 months), 11 patients died, and 3 lost contact, while the others achieved different cytogenetic/molecular responses. The locations of BCR-ABL1 (der(22) vs. non-der(22)) and the karyotype results (complex karyotype vs. noncomplex karyotype) by conventional cytogenetics were not associated with overall survival in this cohort. However, redefining the complexity of chromosomal abnormality by correlating karyotype and FISH findings, CML cases with simple chromosomal abnormalities had a more favorable overall survival than that with complex chromosomal abnormalities. We conclude that insertion-derived BCR-ABL1 fusions often involve complex chromosomal abnormalities which are overlooked by conventional cytogenetics, but can be identified by one or more FISH tests. We also suggest that the traditional cytogenetic response criteria may not apply in these patients, and the complexity of chromosomal abnormalities redefined by correlating karyotype and FISH findings can plays a role in stratifying patients into more suitable risk groups for predicting prognosis. (Word count: 292).