An intestinal proton-dependent peptide transporter located on the lumenal surface of the enterocyte is responsible for the uptake of many orally absorbed beta-lactam antibiotics. Both cephalexin and loracarbef are transported by this mechanism into the human intestinal Caco-2 cell line. Forty-seven analogs of the carbacephalosporin loracarbef and the cephalosporin cephalexin were prepared to evaluate the structural features necessary for uptake by this transport carrier. Compounds were evaluated for their antibacterial activities and for their ability to inhibit 1 mM cephalexin uptake and, subsequently, uptake into Caco-2 cells. Three clinically evaluated orally absorbed carbacephems were taken up by Caco-2 cells, consistent with their excellent bioavailability in humans. Although the carrier preferred the L stereoisomer, these compounds lacked antibacterial activity and were hydrolyzed intracellularly in Caco-2 cells. Compounds modified at the 3 position of cephalexin and loracarbef with a cyclopropyl or a trifluoromethyl group inhibited cephalexin uptake. Analogs with lipophilic groups on the primary amine of the side chain inhibited cephalexin uptake, retained activity against gram-positive bacteria but lost activity against gram-negative bacteria. Substitution of the phenylglycl side chain with phenylacetyl side chains gave similar results. Compounds which lacked an aromatic ring in the side chain inhibited cephalexin uptake but lost all antibacterial activity. Thus, the phenylglycl side chain is not absolutely required for uptake. Different structural features are required for antibacterial activity and for being a substrate of the transporter. Competition studies with cephalexin indicate that human intestinal Caco-2 cells may be a useful model system for initially guiding structure-activity relationships for the rational design of new oral agents.