DNA was analysed by flow cytometry in paraffin-embedded tumour specimens from 58 patients who had undergone resection of pancreatic ductal adenocarcinoma, to see if there was any correlation between DNA analysis and survival. Because of the poor quality of the DNA histogram five cases were excluded from the analysis of DNA-ploidy and 10 from the analysis of the synthesis phase fraction. Aneuploidy was detected in 11 (21%) of the 53 tumours. Neither aneuploidy nor synthesis phase fraction correlated significantly with size, stage or differentiation of the tumours. The survival of the patients with DNA-aneuploid tumours (median 9 months) did not differ significantly from that of the patients with diploid tumours (median 14 months). Four patients survived for 5 years or longer, two of whom had aneuploid tumours. Synthesis phase fraction did not correlate with survival. Using univariate survival analyses, location of the tumour in the head of the pancreas (p = 0.02), a tumour diameter of less than 3.8 cm (p = 0.003), tumour stages I and II (p = 0.03), and a well differentiated tumour (p = 0.04) correlated with favourable survival. In a Cox's regression analysis, however, only the diameter of the tumour had independent prognostic value. We conclude that DNA analysis by flow cytometry is not helpful in selecting patients with resectable pancreatic carcinoma who will benefit from resection.