The presence of complex karotypes with frequent numerical and structural abnormalities has been reported in 20 to 50% of multiple myeloma (MM) patients. This variability is mainly due to the difficulty of conventional cytogenetics to obtain tumor metaphases representative of all possible neoplastic clones in MM. To gain insight into the real incidence of numerical chromosome changes in MM we have studied by fluorescence in situ hybridization technique 15 different human chromosomes, 1, 3, 6, 7, 8, 9, 10, 11, 12, 13, 15, 17, 18, X, and Y, in a series of 52 MM patients. In all cases, the DNA index assessed by a propidium iodide/CD38 double-staining technique with flow cytometry was simultaneously investigated for correlation, with fluorescence in situ hybridization results. Additional aims of this study were 1) to analyze whether the abnormalities detected were common to all plasma cells or were present in only a subpopulation of tumor cells, 2) to explore changes caused by disease progression, and 3) to establish possible associations among the altered chromosomes. Although the overall incidence of numerical abnormalities was 67%, this frequency increased to 80% in the 41 cases in which 7 or more chromosomes were analyzed. Trisomies were significantly more common than monosomies (84% versus 16%). Chromosomes 9 and 15 were the most frequently altered (52% and 48% of cases, respectively), with all of their abnormalities corresponding to trisomies. The most frequent losses involved chromosomes 13 (26%) and X in females (32%). Other common numerical changes corresponded to chromosomes 1 (39%), 11 (37%), 6 (32%), 3 (31%), 18 (29%), 7 (28%), and 17 (22%). By contrast, chromosomes 8(13%), 10(8%), and 12(3%) were rarely altered. DNA aneuploidy by flow cytometry was detected in 67% of patients, and a high degree of correlation was observed between the DNA index obtained by flow cytometry and the chromosome index derived from fluorescence in situ hybridization studies, calculated according to two mathematical formulas (coefficient of correlation of 0.82 and 0.91 when at least 7 or 12 chromosomes were considered, respectively). The frequency of numeric chromosome aberrations was higher in those patients with progressive disease and, interestingly, trisomy of chromosome 8 was exclusively detected in this latter group of patients. Our study shows that, with the exception of chromosome 8, a possible marker of clonal evolution, the numeric chromosome changes are present in nearly all malignant plasma cells (r > 0.84). Finally, frequent associations between chromosomal aberrations were observed (ie, chromosomes 6, 7, 9, and 17; 7 and 15; and 11 and 17). By excluding them, it was found that two triple combinations of chromosome-specific probes, chromosomes 1 and 9 together with either chromosome 13 or 15, could be a useful marker for detection of residual disease, as it permits the identification of most MM patients displaying numerical changes.