Flow cytometric DNA analysis was performed in a total of 203 bone tumors, benign and malignant. In more than 80% of cases the material studied was paraffin-embedded tumor tissue, mainly from the archives of the Bone Tumor Registry of Westphalia in Münster. Compared with ethanol-fixed fresh tumor samples, the variation coefficient in DNA histograms of the stored material was increased by a factor of 1.2-1.5, which means that resolution was decreased and that, in many cases, accurate cell cycle analysis was not feasible. However, the results of cell cycle analysis in bone tumors, even if performed on optimally fixed specimens, have to be evaluated with caution and full reference to the corresponding histological slides, since these histograms are apt to show various superpositions from the inflammatory infiltrate. The assessment of DNA ploidy is unimpaired if, in agreement with most researchers today, deviations smaller than +/- 10% from the diploid standard are still defined as DNA diploid, peridiploid, or pseudodiploid. The coefficient of variation should be kept as low as possible. If it is between 10% and 15%, the near-diploid stemlines with DNA indices of 0.9 or 1.1 may be hard to delineate. On account of the particularly marked regressive changes, the resolution of DNA histograms was most strongly impaired in chondromatous tumors, whereas it was mostly excellent in highly cellular viable tumor tissue, such as that from Ewing's sarcoma or osteoblastoma. On the whole, there was a distinct correlation between DNA ploidy and the biological behavior of bone tumors (Table 8). The highest rates of DNA aneuploidy were found in highly malignant OSs (18/21) and FSs (14/16), thus reflecting their poor prognosis. Of six juxtacortical OSs, three well-differentiated parosteal OSs and two periosteal OSs were DNA diploid, whereas one highly malignant surface OS and five highly malignant extraskeletal OSs, all DNA aneuploid, corresponded fully to the medullary OSs. Judging by preliminary results, adjuvant preoperative chemotherapy (COSS 80/82: Bösing et al. 1987) may reduce the rate of DNA aneuploidy and, consequently, of stem cell heterogeneity in general. A selective destruction of those stemlines that respond particularly to chemotherapy appears probable. In contrast to their high malignancy, Ewing's sarcomas showed an unexpectedly low proportion of DNA aneuploid stemlines (14/24). The comparatively favorable prognosis of MFH of bone is reflected in a lower rate of aneuploidies (2/10), which is also rather low (probably too low) when compared to our own data from soft tissue MFH (9/19).(ABSTRACT TRUNCATED AT 400 WORDS)