The fractional absorption of vitamin B12 (FAB12) was measured by a double-isotope technique specially adapted for children. In six celiac children on a strict gluten-free diet and with a normal small intestinal biopsy, the FAB12 performed in the fasting patient averaged 30% (23-40%). After gluten challenge for a mean of 2 months (range 1-4), when mucosal damage was demonstrated by biopsy, the average fasting FAB12 in these patients decreased to 10% (0-17%) (p less than 0.05). However, when the FAB12 test was repeated by means of stimulation by a B12-free meal 1-3 weeks later, while the patients were still on a diet containing gluten, a significant increase was observed (mean 21%, range 14-27%) (p less than 0.05). In four of the six patients the B12 absorption was further evaluated by repeating the FAB12 test by means of intravenous cholecystokinin (CCK) stimulation (n = 3) or by administration of exocrine pancreas enzyme supplementation (EPES) (n = 2) or cobinamide (n = 1). These tests all showed FAB12 values within the range of the meal-stimulated FAB12. Moreover, in eight gluten-free celiac children with normal biopsies, no difference was found between fasting and meal-stimulated FAB12 values. Therefore, it is likely that the early-onset B12 malabsorption observed in the gluten-challenged celiac child with upper-small-intestinal mucosal damage is in part due to an insufficient stimulation of the exocrine pancreas when using the standard fasting B12 absorption test.(ABSTRACT TRUNCATED AT 250 WORDS)