The healthy relatives of patients with Crohn's disease were previously found to have increased intestinal permeability to polyethylene glycol 400. To determine whether the abnormal permeability is uniquely detectable by polyethylene glycol 400, we studied the intestinal permeability of three new probes (lactulose, rhamnose, and mannitol) in 25 patients with Crohn's disease, 41 of their healthy relatives, and 29 normal controls without a family history of inflammatory bowel disease. Patients with Crohn's disease had increased lactulose permeability when compared with relatives or controls. Lactulose absorption by patients with Crohn's disease was 0.41% +/- 0.07% (mean +/- SE), whereas that of their relatives and unrelated controls was 0.28% +/- 0.03% and 0.26% +/- 0.03%, respectively. There was no significant difference between the relatives and controls, but both groups differed from the patients (p less than 0.05 and p less than 0.025, respectively). The patients' lactulose/rhamnose ratio was 70.5% +/- 9.2% vs. 37.2% +/- 3.3% in relatives and 40.6% +/- 5.7% in unrelated controls (p less than 0.0005 and p less than 0.0025, respectively). The two intermediate-sized probes, rhamnose and mannitol, did not detect permeability differences among the three groups. The inability of lactulose, rhamnose, or mannitol to detect permeability abnormalities in healthy relatives of patients with Crohn's disease suggests that these probes penetrate the intestinal barrier by routes or mechanisms that are different from those of polyethylene glycol 400. Lactulose, in particular, detects permeability changes in patients with intestinal inflammation, and polyethylene glycol 400 is able to detect permeability changes in the health relatives of our patients. These data indicate that permeability may be abnormal as a secondary result of inflammation, or as a result of a primary genetic abnormality.