Twenty eight patients with classical irritable bowel syndrome completed a double blind placebo controlled crossover trial in which they added to their normal diet a daily supplement of either 12 bran biscuits (1 = 1.3 g fibre) or 12 placebo biscuits (1 = 0.23 g fibre). Each biscuit was given for three months in random order with crossover to the alternative biscuit at three months. After the initial three months therapy, there was a significant symptomatic improvement compared with pretreatment in both the bran treated (p less than 0.01) and placebo treated groups (p less than 0.01), but there was no significant difference in symptom scores between these two groups. There was no further improvement in either group after the second three months treatment with the alternative therapy. When crossover data for all 28 subjects were combined, symptoms scores after three months bran therapy and after three months placebo therapy did not differ significantly. Twenty four patients completed three day stool collections in both treatment periods. When the symptomatic response to bran among 15 subjects in whom stool weights rose on bran was compared with that among nine subjects whose stool weights were static or fell on the bran, it was shown that symptomatic improvement was independent of an increase in stool weight. These data suggest that in irritable bowel syndrome, especially that associated with abdominal pain, the beneficial effects of bran are due to a placebo response which is independent of an increase in stool weight.