The paper discusses the principal evidence that supports the concept that cell and tissue injury in infectious and post-infectious and inflammatory sequelae might involve a deleterious synergistic interaction among microbial- and host-derived pro-inflammatory agonists. Experimental models had proposed that a rapid cell and tissue injury might be induced by combinations among subtoxic amounts of three major groups of agonists generated both by microorganisms and by the host's own defense systems. These include: (1) oxidants: Superoxide, H(2)O(2), OH', oxidants generated by xanthine-xanthine-oxidase, ROO; HOC1, NO, OONO'-, (2) the membrane-injuring and perforating agents, microbial hemolysins, phospholipases A(2) and C, lysophosphatides, bactericidal cationic proteins, fatty acids, bile salts and the attack complex of complement a, certain xenobics and (3) the highly cationic proteinases, elastase and cathepsin G, as well as collagenase, plasmin, trypsin and a variety of microbial proteinases. Cell killing by combinations among the various agonists also results in the release of membrane-associated arachidonate and metabolites. Cell damage might be further enhanced by certain cytokines either acting directly on targets or through their capacity to prime phagocytes to generate excessive amounts of oxidants. The microbial cell wall components, lipoteichoic acid (LTA), lipopolysaccharides (LPS) and peptidoglycan (PPG), released following bacteriolysis, induced either by cationic proteins from neutrophils and eosinophils or by beta lactam antibiotics, are potent activators of macrophages which can release oxidants, cytolytic cytokines and NO. The microbial cell wall components can also activate the cascades of coagulation, complement and fibrinolysis. All these cascades might further synergize with microbial toxins and metabolites and with phagocyte-derived agonsits to amplify tissue damage and to induce septic shock, multiple organ failure, 'flesh-eating' syndromes, etc. The long persistence of non-biodegradable bacterial cell wall components within activated macrophages in granulomatous inflammation might be the result of the inactivation by oxidants and proteinases of bacterial autolytic wall enzymes (muramidases). The unsuccessful attempts in recent clinical trials to prevent septic shock by the administration of single antagonists is disconcerting. It does suggest however that, since tissue damage in post-infectious syndromes is most probably the end result of synergistic interactions among a multiplicity of agents, only agents which might depress bacteriolysis in vivo and 'cocktails' of appropriate antagonists, but not single antagonists, if administered at the early phases of infection especially to patients at high risk, might help to control the development of post-infectious syndromes. However, the use of adequate predictive markers for sepsis and other post-infectious complications is highly desirable. Although it is conceivable that anti-inflammatory strategies might also be counter-productive as they might...