Cell proliferation in the small-intestinal crypts of rodents has been intensively investigated and lends itself to the deployment of techniques which are inapplicable in man. In particular there are ethical and economic objections to methods involving the use of tritiated thymidine in vivo for specific labelling of DNA, while the validity of in vitro studies using organ culture is uncertain. It is possible, nevertheless to construct a profile of the size and cytokinetic status of the mucosal crypts by analysis of serial sections prepared from well orientated routine diagnostic biopsy specimens. Such studies can provide a measure of mean total crypt-cell population, and by studying the distribution of mitoses in the crypts relative measurements can be obtained of proliferation and maturation compartment sizes, and of crypt cell production rate (CCPR). These parameters have been compared in 62 patients with 'flat' avillous coeliac mucosae and in 85 patients with normal villous mucosae. A heterogeneous group of 47 patients with lesser degrees of abnormality (convoluted mucosae) were similarly studied. In addition, estimates of cell cycle times were obtained in a small group of patients with normal, convoluted and 'flat' mucosae by taking biopsies before and after the administration of the metaphase-arresting agent vincristine. 'Flat' coeliac mucosae show a threefold increase in the size of the proliferation compartment compared with normal and the cell cycle time is approximately halved leading to a net sixfold increase in CCPR. This is the basis of the change in mucosal morphology and presumably represents a compensatory reaction to the gluten-induced increase in loss of enterocytes from the mucosal surface. Convoluted mucosae occupy an intermediate position in terms of the parameters studied and should be regarded as stages in a continuum of adaptive change.