It has generally been assumed that resistance of animals to the pyrogenic and lethal effects of bacterial endotoxin is largely dependent on the phagocytic capacity of the reticulo-endothelial system. Thus it has been shown that stimulation of the reticulo-endothelial system with non-lethal doses of bacterial lipopolysaccharide or endotoxin results in protection of the animals against a subsequent lethal dose of endotoxin1. Again, numerous workers have demonstrated that a ‘blockade’ of the reticulo-endothelial system by colloidal particles, for example saccharated iron oxide, thorotrast and trypan blue, increases the sensitivity of the animals to the lethal effects of these substances2. Recently, however, it has been shown that in certain circumstances it is possible for mice to possess an increased phagocytic activity, as measured by their ability to clear colloidal carbon, and yet be highly sensitive to the effects of endotoxin. This can be seen during the course of an intracellular infection3, or following the administration of zymosan4. The apparent contradiction of these results has yet to be resolved, and it is not unlikely that such resolution may lead to an elucidation of the mode of action of the endotoxins.