An in vitro methodology which mimics in vivo human upper gastrointestinal transit was developed. The transit tolerance of potentially probiotic Lactobacillus and Bifidobacterium species was determined by exposing washed cell suspensions at 37 degrees C to a simulated gastric juice (pH 2.0), containing pepsin (0.3% w/v) and sodium chloride (0.5% w/v), and a simulated small intestinal juice (pH 8.0), containing pancreatin USP (1 g l-1) and sodium chloride (5 g l-1), and monitoring changes in total viable count periodically. The methodology was also employed to determine the effect of adding milk proteins (1 g l-1), hog gastric mucin (1 g l-1) and soyabean trypsinchymotrypsin inhibitor [SBTCI] (1 g l-1) on transit tolerance. The majority (14 of 15) of isolates lost > 90% viability during simulated gastric transit. Only one isolate, Lactobacillus fermentum KLD, was considered intrinsically resistant. The addition of milk proteins, singly and in combination, generally improved gastric transit tolerance. In this regard, two isolates, Lact. casei 212.3 and Bifidobacterium infantis 25962, exhibited 100% gastric transit tolerance in the presence of milk proteins. In general, the addition of hog gastric mucin did not influence simulated gastric transit tolerance of lactobacilli but tended to increase that of bifidobacteria. However, it increased that of Lact. casei 242 and Lact. salivarius 43338 but diminished that of B. bifidum 2715 and B. animalis Bo. Selected bile salts-resistant isolates were intrinsically tolerant to simulated small intestinal transit. Only Lact. casei F19 and B. adolescentis 15703T showed significant reduction in viability after 240 min. In general, the addition of milk proteins and SBTCI did not affect simulated small intestinal transit tolerance. However, they significantly improved the intrinsic resistance of Lact. casei F19 but diminished that of B. breve 15700T. It is concluded that, whereas the majority of bile salts-resistant lactobacilli and bifidobacteria may be intrinsically sensitive to gastric transit, they are intrinsically resistant to small intestinal transit. In addition, it is postulated that milk proteins and mucin may function as both buffering agents and inhibitors of digestive protease activity in vivo, thereby protecting ingested bacterial strains during upper gastrointestinal transit.