Results presented here and by others indicate that the release of colicins from producing cells can be uncoupled from the decline in culture turbidity which usually occurs within 2-3 h after the induction of colicin synthesis. This excludes lysis as a necessary event in colicin release. Conversely, the failure to dissociate colicin release from the normally simultaneous release of a specific subset of soluble proteins argues against the idea of a specific colicin secretion system sensu-stricto. Rather, colicin release appears to be a consequence of semi-specific leakage resulting from an alteration of the permeability properties of the cell envelope. This alteration is caused by the 'lysis protein' known to be encoded by most multiple copy number Col plasmids. The finding that the expression of the lysis gene of plasmid ColE2 renders the cells exquisitely sensitive to lysozyme demonstrates that the permeability of the outer membrane must indeed be altered. Evidence is presented that this alteration could be due at least in part to the activation of the detergent-resistant phospholipase A (pldA product). Lysophosphatidylethanolamine, a product of the action of phospholipase on phosphatidylethanolamine, is a membrane perturbant which could alter the permeability properties of the envelope and allow some proteins such as colicin to leak out of the cell.