To evaluate factors influencing the short-chain fatty acid (SCFA) concentrations in stools, three different experiments were performed: faecal concentrations of SCFA at defecation were determined by gas liquid chromatography in nine healthy volunteers on a free diet. SCFAs were 114 +/- 15.0 mmol/l (means +/- SD). The coefficient of variation (CV) of the assay was 4-15%, the intraindividual CV 12-33%, and the interindividual CV 11-29%. On incubation of faeces at 37 degrees C concentrations of SCFA doubled in 6 h and rose fourfold in 72 h. In three volunteers the experiments were extended by adding ispaghula husk or lactulose to the diet for two 14-day periods each; no change in faecal SCFA concentrations was seen, either at defecation or after incubation. When ispaghula husk or lactulose was added to faeces in an in vitro incubation system, the concentrations of SCFA were five times higher than those of controls. We conclude that instant handling of faeces is essential for determinations of SCFA concentrations to obtain interpretable and comparable results; that determination of total SCFA output is of limited value; that addition of fibre to the diet does not influence faecal SCFA concentrations; and that the capacity for SCFA production in faeces is large provided a sufficient amount of substrate is available.