Flow-cytometric DNA analysis of human tumors using paraffin-embedded tissue samples is becoming an increasingly popular method of determining ploidy and proliferative rate. Particularly with hematopoietic/lymphoid proliferations, little is known about how these data compare with data from fresh suspension studies, or how B5-fixed tissues compare with those fixed in formalin. For these reasons, flow-cytometric DNA ploidy and cell cycle analysis was performed on 16 hyperplastic tonsils and 28 lymphoid/hematopoietic neoplasms using fresh (or fresh-frozen) cell suspensions (FR), B5-fixed paraffin-embedded (B5), and formalin-fixed paraffin-embedded (FO) tissue. Ploidy analysis showed all tonsil specimens to be diploid regardless of preparative method; however, for the neoplastic cases the FO and B5 preparations agreed with the FR preparations in 50% and 61% of the cases, respectively. Complete agreement between all three tissue preparation methods was present in only 36% of the neoplastic cases. Percent S or S + G2M phase fractions showed relatively poor correlation between the three preparative methods, with the best correlations found between the FR and B5 samples (S phase, r = 0.44; S + G2M, r = 0.43). In conclusion, while potentially useful data can be obtained from flow-cytometric DNA analysis of fixed, paraffin-embedded lymphoid/hematopoietic tissues, the specific limitations of such analyses must be recognized. Ploidy and proliferative phase data from fixed tissue preparations are not equivalent to information obtained from fresh suspension studies. B5-fixed tissue provides an acceptable alternative to formalin-fixed tissue for such analyses.