Ploidy and cell proliferation determined by flow cytometry were assessed on colorectal cancers from patients admitted to two Italian cancer research centres. A total of 181 patients were followed prospectively for 4 years at the Istituto Regina Elena (IRE) of Rome and at the Istituto Nazionale Tumori (INT) of Milan. Fresh (at the IRE) or frozen (at the INT) tumour material and similar procedures were used for subsequent sample preparation. Similar frequencies of aneuploid tumours (63% vs 66%) and superimposable median DNA indices (1.6) were observed for the two case series. In both series, DNA ploidy was generally unrelated to clinico-pathological factors, except for a higher frequency of aneuploid tumours in Dukes' D (88%) than in Dukes' A stage (33%) in the IRE experience. DNA ploidy was a weak prognostic indicator at 3 years but not at 4 years in the IRE case series, and it never exhibited a clinical relevance in the INT experience. Conversely, multiploidy was an indicator of worse relapse-free and overall survival at 4 years in the IRE and INT case series.