The principal substrates for colonic bacterial growth are dietary carbohydrates which have escaped digestion in the upper gastrointestinal tract. These may be starches, dietary fibres, other non-absorbable sugars, sugar alcohols and oligosaccharides. In the large intestine, saccharolytic bacteria are able to metabolize carbohydrates for increased energy and growth with short-chain fatty acids (SCFA) and a variety of other metabolites, such as the electron-sink products lactate, pyruvate, ethanol, H, and succinate, being produced. The majority of human large intestinal micro-organisms, have a strictly anaerobic metabolism, whilst numbers of facultative anaerobes are many orders of magnitude lower than those of the obligate anaerobes. Of the culturable flora, numerically predominant anaerobes are Gram-negative rods belonging to the genus Bacteroides. Other groups which have hitherto been identified as quantitatively significant include bifidobacteria, clostridia, eubacteria, lactobacilli, Gram-positive cocci, coliforms, methanogens and dissimilatory sulphatereducing bacteria. Generally, the various components of the large intestinal microbiota may be considered as exerting either pathogenic effects or they may have potential healthpromoting values. Bifidobacteria and lactobacilli are considered to belong to the latter group. Bacteria in the colon respond largely to the available fermentable substrate, and there is currently some interest in the use of diet to specifically increase groups perceived as health promoting. Non-digestible oligosaccharides seem to have this (prebiotic) potential. In particular, those that contain fructose are well fermented by bifidobacteria such that their numbers become predominant in the faeces of volunteers fed on the oligosaccharides. Whilst dietary modification of the microflora composition is therefore possible, it remains to be determined whether such effects have any positive health attributes. New molecular-based methodologies for the improved detection of gut bacteria, including species that are non-culturable by conventional methodology, will give increased precision for understanding the effects of diet on the colonic microbiota composition.