Intestinal permeability was evaluated in 17 healthy young children (control group), aged 4.3 +/- 1.9 years (mean +/- SD), weight 15.4 +/- 5.5 kg and in 11 infants with food allergy (allergic group, aged 2.3 +/- 1.6 years, weight 9.5 +/- 1.5 kg). In the two groups, the urinary excretion of two non-metabolized sugars (mannitol, lactulose) was measured by gas-chromatography after oral administration. Measurement of urinary content of these sugars 5 hours after administration in the control group showed a clearance of 20.2 +/- 8.6% for mannitol and 0.44 +/- 0.18% for lactulose. The lactulose/mannitol ratio in the same group was equal to 0.023 +/- 0.008. When compared to the control group, the 11 patients of the allergic group presented a normal mannitol urinary excretion (16.5 +/- 13.4%, p = NS, Student's t-test) and an increase in the lactulose excretion (1.36 +/- 0.92%, p < 0.001). Moreover, the allergic group showed a lactulose/mannitol ratio that was significantly different (0.105 +/- 0.071, p < 0.001). The intestinal permeability test for the diagnosis of food allergies seems to be a sensitive and non-invasive test that is well suited to the paediatric practice.