In vitro experiments have shown that fermentation of carbohydrates prevents accumulation of nitrogen in the colon. Variable results have been obtained on modulation of dietary intakes in vivo. Lactose[15N,15N]-labeled ureide has been proposed as a tool to study colonic nitrogen metabolism. However, on oral administration of the marker, different urinary excretion patterns of the 15N label have been found. In this study, 50 mg lactose[15N,15N]ureide was directly instilled in the colon through an orocecal tube to investigate the colonic handling of this molecule in a direct way. In basal conditions, 42% (range, 37-48%) of labeled nitrogen administered as lactose[15N,15N]ureide was retrieved in urine after 72 h. A substantial variability in total urinary excretion of the label was found, but the urinary excretion pattern of the label was similar in all volunteers. When inulin, a fermentable carbohydrate, was administered together with the labeled marker, a significant decrease in urinary excretion of 15N after 72 h was found, to 29% (range, 23-34%). The effect of a smaller dose of inulin (250 mg) on colonic handling of lactose[15N,15N]ureide (50 mg), was investigated in another group of volunteers, and this time, fecal excretion of the marker was also evaluated. The results seem to indicate that fermentation of inulin causes an increased fecal excretion of the marker, thereby reducing urinary excretion but not retention in the human nitrogen pool. This instillation study shows that lactose[15N,15N]ureide is a tool with good properties to investigate the effect of different types of carbohydrates on nitrogen metabolism in the proximal colon in vivo.