The aim of this study was to determine the value of the lactulose mannitol intestinal permeability test in screening the general adult population for unrecognized enteropathy and latent coeliac disease. Subjects with positive serology (identified by screening carried out by the Belfast MONICA Project) along with controls were followed-up after 3 years and classified as having transient serology, persistent serology or coeliac disease. A 5-h urine collection was performed following the ingestion of 5 g lactulose, 2 g mannitol and glucose as an osmotic filler. Urinary concentrations of lactulose and mannitol were measured by enzymatic analysis. Percentage lactulose excretion (%LE) (0.94 versus 0.31, P<0.001) and lactulose mannitol excretion ratio (LMER) (0.12 versus 0.02, P<0.001) were significantly higher in screening-detected coeliac disease subjects compared with MONICA controls. The sensitivity of the permeability test was 87% in the screening situation compared with 81% in the clinical situation. In subjects with persistent and transient serology the LMER did not differ significantly from that of controls. The lactulose-mannitol test is a useful test for screening the general adult population for coeliac disease. Subjects with persistent and transient serology did not differ from MONICA controls and are unlikely to have latent coeliac disease.