The ability to retain viscous fluid in the standing position was tested in 22 patients with fecal incontinence, 11 patients with constipation, and 26 control subjects. Viscous fluid was introduced into the rectum in increments of 50 ml. The examination was stopped when the patient complained of discomfort or the viscous fluid leaked. Eighteen of 22 patients with fecal incontinence leaked fluid, while none of the control subjects and only four of the constipated patients did so. Patients with fecal incontinence retained significantly less viscous fluid than did control subjects, whereas no difference was found between patients with constipation and control subjects. Rectal sensation from distention with air was tested in the patients as well as in the control group. The following volumes and pressures at each sensation were measured: 1) earliest defecation urge (EDU), 2) constant defecation urge (CDU), and 3) maximum tolerable volume (MTV). Patients with fecal incontinence had lower volumes than control subjects at all sensations, while patients with constipation had higher volumes at earliest defecation urge and at constant defecation urge. Rectal compliance was higher in patients with fecal incontinence than in control subjects, whereas patients with constipation did not differ from control subjects. Regression analysis showed a linear relationship between viscous fluid retention and the maximum tolerable volume and also between viscous fluid retention and rectal compliance. No difference in the ability to retain viscous fluid between male and female control subjects was found; regression analysis of viscous fluid retention in relation to age revealed decreasing volumes with increasing age. Day-to-day variation of the ability to retain viscous fluid was tested in eight persons, and reproducibility was found to be good.