The functional integrity of the small bowel is impaired in coeliac disease. Intestinal permeability, as measured by the sugar absorption test probably reflects this phenomenon. In the sugar absorption test a solution of lactulose and mannitol was given to the fasting patient and the lactulose/mannitol ratio measured in urine collected over a period of five hours. The sugar absorption test was performed in nine patients with coeliac disease with an abnormal jejunum on histological examination, 10 relatives of patients with coeliac disease with aspecific symptoms but no villous atrophy, six patients with aspecific gastrointestinal symptoms but no villous atrophy, and 22 healthy controls to determine whether functional integrity is different in these groups. The lactulose/mannitol ratio (mean (SEM) is significantly higher in both coeliac disease (0.243 (0.034), p < 0.0001)) and relatives of patients with coeliac disease (0.158 (0.040), p < 0.005)) v both healthy controls (0.043 (0.006)) and patients with aspecific gastrointestinal symptoms (0.040 (0.011)). The lactulose/mannitol ratio in relatives of coeliac disease patients was significantly lower than in the coeliac disease patient group (p = 0.04). The lactulose/mannitol ratio was the same in healthy controls and patients with aspecific gastrointestinal symptoms. It is concluded that the sugar absorption test is a sensitive test that distinguishes between patients with coeliac disease and healthy controls. The explanation for the increased permeability in relatives of patients with coeliac disease is uncertain. Increased intestinal permeability may be related to constitutional factors in people susceptible to coeliac disease and may detect latent coeliac disease. The sugar absorption test may therefore be helpful in family studies of coeliac disease.