This study was undertaken to assess the value of flow cytometric measurements of total p53 protein content and proliferation indices derived from in vivo halogenated pyrimidine labelling. Two series of colorectal cancer specimens were studied for which clinical outcome data were recorded. A series of 84 archival, ethanol-fixed, bromodeoxyuridine (BrdUrd) labelled colorectal tumours were analysed by flow cytometry for their total and cell cycle phase p53 protein content using the pAb1801 monoclonal antibody. A second series of 33 freshly obtained tumours was used for assay evaluation and for comparison with the archival material. In the archival series (n=84), the median p53-pAb1801 LI was 81.9% (range: 11.1-99.8%). In only three tumours could significant amounts of p53 protein not be detected. The median phase specific p53-pAb1801 LI in G0/G1 was 71.6%, in S was 95.5%, and in G2/M was 98.5%. In the series of fresh tumours (n=33), the median p53-pAb1801 labelling index (LI) was 94.6% (range: 17.9-99.9%). Only two tumours failed to express significant amounts of p53 protein. There was no significant difference in the generally high levels of p53 protein content between the fresh and archival series. Life-table analysis of the patients in the archival series failed to demonstrate a statistical difference in life expectancy in relation to Dukes' stage when tumours were stratified by the median total p53 labelling index. In this study, p53 content and proliferative indices measured by flow cytometry do not have independent predictive value over Dukes' grading in determining the outcome of colorectal cancer. Flow cytometry is confirmed as a practical tool for multi-parametric and cell cycle analysis of oncoprotein expression in human tumour biopsies.