Objective: Comparative genomic hybridization (CGH) has been established as an informative technique in genetic analysis. However, differences in the ratio of hybridization intensities were reported for particular chromosomes, which may affect CGH results. The aim of this study was to define these differences in more detail. For this purpose, CGH results of 70 samples of bone marrow cells (BMC) with normal karyotype in conventional cytogenetics (CC) were evaluated using seven different reference DNAs and two different DNA labeling systems. Methods and Results: CGH using fluorochrome-conjugated nucleotides for DNA labeling indicated signal deviations in 21/70 BMC samples. Deviations affected chromosomes 1 (n = 21), 2 (n = 11), 4 (n = 11), 5 (n = 9), 6 (n = 7), 7 (n = 2), 8 (n = 2), 12 (n = 5), 13 (n = 15), 14 (n = 1), 16 (n = 17), 17 (n = 11), 19 (n = 21), 20 (n = 12), and/or 22 (n = 17). None of the imbalances were confirmed by fluorescence in situ hybridization (FISH). Using digoxigenin and biotin-conjugated nucleotides in exemplary cases (n = 5) led to the disappearance of the signal deviations. Repeated CGH experiments using seven different reference DNAs showed remarkable variations in the signal deviations. Conclusion: Hybridization differences depend not only on the hapten or fluorochrome-labeled nucleotides used for DNA labeling, but also on the reference DNA chosen. Therefore, close control of CGH experiments is mandatory, and additional techniques such as FISH should be performed to confirm the results obtained by CGH.