We applied a dual-color interphase in situ fluorescence hybridization (I-FISH) technique using centromeric probes specific to chromosomes 7 and 8 on 20 control samples in order to define the statistical model best suited to determine cutoff values for detection of small abnormal clones. We found that the Poisson model is a more appropriate approach than a Gaussian model. Then, based on the analysis of 91 samples from 80 patients with myelocytic malignant hemopathies and either clonal or nonclonal -7 or +8 as determined with conventional cytogenetics (CC), we compared the respective power of I-FISH and CC for detection of aneuploidy, with special emphasis on the potential contribution of I-FISH as a complement to CC in the case of small abnormal clones. The I-FISH results were positive in samples with clonal -7 or +8 according to CC analysis. Whereas I-FISH was negative in samples with nonclonal -7 according to CC, thus confirming the reliability of the criteria used to define the clonality of -7; the situation was different with nonclonal +8. I-FISH revealed the clonality of +8 in most samples with single-cell +8. In several cases, however, the unquestionable clonal nature of +8, as evidenced during follow-up, could not be established with either CC or I-FISH according to accepted criteria. Our data suggest that, in case of a single metaphase with +8, the general rule should be amended and the single-cell +8 should be considered and reported as potentially clonal.