Symptom scores, stool data, and the transit of a standard, solid meal were measured in 28 patients with irritable bowel syndrome (IBS) during baseline conditions and after five weeks of treatment with placebo and loperamide, given as a flexible dosage regime in the form of a double-blind, cross-over trial. All patients had undergone a comprehensive series of diagnostic investigations and had failed to respond to dietary supplementation with coarse wheat bran (10-30 g daily). Loperamide treatment accelerated gastric emptying, compared with placebo (1.2 +/- 0.1 vs 1.5 +/- 0.1 hr; P less than 0.001) and delayed both small bowel (6.2 +/- 0.3 vs 4.3 +/- 0.3 hr; P less than 0.001) and whole gut transit (56 +/- 5 vs 42 +/- 4 hr; P less than 0.01). Eighteen patients said they felt better taking loperamide compared with placebo and, at follow up, 15 of these patients remained satisfied with the effects of the drug. Most symptoms improved significantly on placebo compared with the baseline period, but three of these [diarrhea (P less than 0.01), urgency (P less than 0.01) and borborygmi (P less than 0.05)] showed a further significant improvement on loperamide. Improvement in diarrhea was not associated with any change in stool weight but was associated with reductions in stool frequency (P less than 0.001), passage of unformed stools (P less than 0.01), and incidence of urgency (P less than 0.001). Urgency was the only symptom that was significantly more common in the success group, compared with the group who did not feel better on loperamide.