OBJECTIVE To explore the distinction between true clinical constipation and the subjective complaint of constipation in elderly people and to review the pathophysiology, symptoms, diagnosis, causes, and treatment. METHODS A computer-assisted and manual search of the English language literature using MEDLINE 1966-1991, Index Medicus 1988-1992, reference lists of selected articles, and relevant textbooks. METHODS Studies that provide information on lower bowel function and laxative and enema use in the elderly subjects were reviewed. Article selection was not limited by study design. METHODS Relevant data were abstracted from the results of physiological, cohort and case-control studies, and clinical trials. The text discusses the methodological strengths and flaws of these studies and excludes management approaches formulated from uncontrolled clinical observation. RESULTS Constipation of the elderly is not well defined in the current literature. Self-reported constipation and laxative use increase with age, while a similar escalation in true clinical constipation is not shown. Physiological changes in the lower bowel predisposing toward constipation do not occur with normal aging. Patient selection criteria for studies examining the pathophysiology of constipation differ in their definition of constipation and their inclusion of coexisting chronic illness. Nevertheless, there is consistent evidence for prolonged transit through the sigmoid colon and rectum, especially in frail elderly patients, and reduced rectal tone with impaired sensation, particularly in patients with rectal impaction. Few studies rigorously examine "risk factors" and non-pharmacological interventions in constipation. The results of most laxative trials require cautious interpretation because of inclusion of patients without diagnostically proven constipation, use of combined laxative preparations, and unreliable outcome measures. Certain laxative agents however appear more appropriate for use in elderly people. CONCLUSIONS Although the subjective complaint of constipation and habitual laxative use increase with age, the epidemiological data suggest that true clinical constipation does not. Physiological changes predisposing toward constipation are not an inevitable consequence of aging, but appear to be specific to the condition. The available data do not confirm many suspected "risk factors" nor the benefits of commonly used non-pharmacological and pharmacological treatments, but they do provide enough information to formulate a practical approach to constipation in elderly persons.