With the use of fresh cell suspensions, the DNA ploidy of 11 benign and 27 malignant spindle cell soft-tissue lesions was determined by flow cytometry (37 cases) and/or image cytometry (35 cases). Of the 27 malignant lesions, 10 were low- or intermediate-grade sarcomas, and 17 were high-grade sarcomas. In the malignant lesions, the DNA ploidy was correlated with the histologic grade and the clinical outcome. Of the 11 benign lesions, six had a diploid DNA ploidy pattern, and five were nondiploid by either flow cytometry or image cytometry. All benign cases had a favorable outocome regardless of ploidy. Eight of the 10 low- or intermediate-grade malignant lesions were diploid, whereas two were nondiploid. Of the 17 high-grade sarcomas, 15 were nondiploid by either method of measurement and nine by both. Of the 10 diploid malignant tumors, only one had an unfavorable outcome (malignant mesothelioma with biopsy only), whereas of the 17 malignant lesions that were nondiploid, five had no evidence of recurrent disease, three cases recurred, eight patients died of disease, and one patient died of an unrelated cause. We concluded that (1) ploidy does not differentiate benign from malignant spindle cell soft-tissue tumors, (2) the nondiploid DNA pattern is more common in high-grade than in low- or intermediate-grade sarcomas, and (3) there is a statistically significant relationship between nondiploidy and a worse clinical outcome in sarcomas.