Permeability tests are widely used to investigate the pathogenesis of various gastrointestinal diseases including coeliac disease, infectious diarrhoea, and inflammatory bowel disease. In Crohn's disease they are used as activity parameters by some investigators. Lack of standardisation, however, makes it very difficult to compare data reported in different studies. The aim of this study was to gather permeation data in well controlled test conditions to standardise the methods. Nine healthy volunteers each received five consecutive permeability tests by mouth using polyethylene glycol-400 (PEG-400) and 51Cr-EDTA as probe molecules. The probes were dissolved in water, a glucose solution, a starch solution, a hyperosmolar lactulose-mannitol solution, and a liquid meal. A significantly decreased permeation for both probes was found when given with the hyperosmolar solution. The 51Cr-EDTA permeation was also decreased with water. The permeability index, 51Cr-EDTA/PEG-400, corrected for influencing factors, confirmed that the lactulose-mannitol solution and plain water yield lower values of macro-molecule permeation than starch, glucose or liquid meal. Hyperosmolarity was clearly accompanied by a decrease in permeability probably caused by reversed solvent drag. Interindividual variability of probe permeation and permeability index is very low with a standard liquid meal. It is proposed that for permeability studies a standard liquid meal is always used.