Cell membranes, including their individual components like membrane-bound proteins and particular lipids, attract a growing attention as very perspective pharmacological targets. Rational design of new efficient and selective compounds modulating activity of biomembranes, requires atomic-scale information on their spatial structure and dynamics under different conditions. Because such details resist easy experimental characterization, important insight can be gained via computer simulations.We present the results of structural/dynamic computational studies of membrane proteins and peptides with diverse fold, mode of membrane binding, and biological activities: antimicrobial and cell-penetrating peptides, cardiotoxins from snake venom, transmembrane domains of receptor tyrosine kinases. The computational approach combines Monte Carlo simulations in implicit membranes, molecular dynamics in full-atom lipid bilayers, and molecular hydrophobicity potential analysis. The predictive power of the method was proven via testing against high-resolution experimental data.Despite different structure and mechanism of membrane permeation, in all cases the polypeptide-membrane recognition reveals a prominent “self-adapting” character. Namely, the membrane active agents employ a wide arsenal of structural/dynamic tools in order to insert into lipid bilayer and to accomplish their function. Importantly, lipid bilayer of biological membranes plays essential role in the recognition/binding events. In particular, the membrane surface reveals highly dynamic lateral heterogeneities (clusters), which differ in their packing and hydrophobic properties from the bulk lipids. Such a mosaic nature of membranes is tuned in a wide range by the chemical nature and relative content of lipids, presence of ions, etc. This makes possible mutual adaptation of the two amphiphatic systems (peptide and membrane). Such a diversity of the factors important for polepeptide-bilayer interactions assures their efficient and robust binding to cell membranes. Understanding of such effects creates a basis for rational design of new physiologically active molecules and/or artificial membranes with predefined properties.