Amongst the various claimed beneficial effects of pro- and prebiotics for the human host, it has been hypothesised that functional foods are able to suppress the generation and accumulation of toxic fermentation metabolites (NH3, p-cresol). Direct evidence supporting this hypothesis is lacking mainly because of the unavailability of reliable biomarkers. Preliminary data indicate that lactose-[15N]ureide and [2H4]tyrosine may be potential biomarker candidates. The aim of the present study was to evaluate the effect of pro- and prebiotics on the colonic fate of these biomarkers in a randomised, placebo-controlled, cross-over study with nineteen healthy volunteers. At the start of the study and at the end of each 2-week study period, during which they were administered either a probiotic (n 10; 6.5 x 10(9) Lactobacillus casei Shirota cells twice daily) or a prebiotic (n 9; lactulose 10 g twice daily), the volunteers consumed a test meal containing the two biomarkers. Urine was collected during 48 h. Results were expressed as percentage of the administered dose. As compared with the placebo, the decrease in the percentage dose of p-[2H4]cresol in the 24-48 h urine fraction was significantly higher after probiotic intake (P=0.042). Similar changes were observed for the 15N tracer (P=0.016). After prebiotic intake, a significantly higher decrease in the percentage dose of p-[2H4]cresol (P=0.005) and 15N tracer (P=0.029) was found in the 0-24 h urine collection. The present results demonstrate that suppression of the generation and accumulation of potentially toxic fermentation metabolites by pro- and prebiotics can reliably be monitored in vivo by the use of stable isotope-labelled biomarkers.