Membrane-disrupting bolaphiles I [HO(CH2CH2O)6CO(CH2)14CO2(CH2CH2O)6H] and III [HO(CH2CH2O)6-CO(CH2)6CH = CH(CH2)6CO2(CH2CH2O)6H], and their polymeric counterparts II (-[CO(CH2)14CO2(CH2CH2O)13]4.8-) and IV (-[CO(CH2)6CH = CH(CH2)6CO2(CH2CH2O)13]5.7-), show high selectivity toward lipid bilayers of varying cholesterol content. In the absence of cholesterol, these surfactants are effective in inducing the release of 5(6)-carboxyfluorescein, entrapped within liposomes made from both egg PC and 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC). The ability of I-IV to disrupt POPC bilayers that contain substantial amounts of cholesterol (> 33 mol %), however, is significantly reduced. In contrast, Triton X-100 and a single-chain analogue of II (i.e., CH3(CH2)6CO2(CH2CH2O)6H, V) are relatively insensitive to the presence of cholesterol. Similar selectivity has been observed with biological targets, i.e., human erythrocytes and a human bacterium (Proteus mirabilis). These results provide the first clear evidence that modest and definable differences in membrane composition and packing can lead to large differences in lability, and that synthetic agents can be created which exploit such differences. The implications of these findings to the development of membrane-disrupting antimicrobial agents are briefly discussed.