Clinical and experimental studies have shown that the carcinogenic process of colorectal mucosa is linked to the development of proliferative abnormalities which precede the occurrence of morphological abnormalities such as epithelial dysplasia. In humans, proliferative defects have been demonstrated in the normal rectal mucosa of population groups at risk for colon cancer. Several techniques are available to study cell kinetics in the gastrointestinal mucosa. Each explores different aspects of cell proliferation. We have attempted to evaluate the correlation between various techniques in the normal rectal mucosa of high-risk patients. We found a good correlation between bromodeoxyuridine (BrdU) labeling index and the mucosal activity of the enzyme ornithine decarboxylase, and between tritiated thymidine and the percentage of cells in the S phase of the proliferative cycle as determined by flow cytometry. However, these correlations concern only labeling indices, which can be influenced by physiological events, such as active inflammation or increased cell loss. It may not be the most reliable proliferative marker of cancer risk. Moreover, methods using cells isolated from homogenized tissue do not allow us to evaluate the pool of cells which are examined nor the distribution of proliferating cells within the tissue. For example, an inflamed mucosal specimen is highly infiltrated with inflammatory cells which can interfere with the measurement of epithelial cell proliferation. Nevertheless, the labeling index may be normal even in patients at high risk. For these reasons, we think that the most reliable methods are those using tissue culture, such as tritiated thymidine or BrdU uptake and proliferating cell nuclear antigen (PCNA) immunostaining.