According to the theory of Fredericq (1949) and Nomura (1964), colicins are attached by specific receptor sites in the cell walls of sensitive bacteria, which mediate their inhibitive effects. During last years, a great variety of experimental data have been accumulated, some of which cannot be easily interpreted in terms of this theory. There exist considerable discrepancies concerning the chemical nature and molecular weight of isolated receptors. The attachment of a colicin onto its receptor need not be irreversible. The inhibition of numerous membrane-associated functions in colicin-tolerant mutants suggests their pleiotropic deletion nature. The difference between colicin resistance and colicin tolerance does not seem to be clear-cut. Cells of stable L-forms of protoplast type, completely devoid of their walls, retain in most cases the same patterns of sensitivity to colicins as rods of the same strains. Experimental changes in the relationship between the cell wall and the cytoplasmic membrane decrease colicin sensitivity of the cells. Colicin E3 has been found to be a specific endoribonuclease, able to cleave a terminal fragment from the 16 S rRNA also in isolated ribosomes in vitro: not only in ribosomes from sensitive bacteria, but also in those from resistant ones and from eukaryotic cells. A destabilization of the DNA helix was induced by colicin E2 in vitro as in vivo. It seems that there exist two distinct types of colicin receptors with different functions: those in the cell wall, and those in the cytoplasmic membrane. Only the contact of colicins with the latter ones is biologically effective and starts both stages of their inhibitive effect: the reversible and the irreversible ones.