Some factors affecting the use of chromogenic substrates with Limulus lysate for assaying bacterial endotoxins in blood have been assessed. It was found that endogenous amidases, which degrade the substrate, could be inactivated by heating serum at 60 degrees C for 15 min. Endotoxin was found not to be removed from serum during clotting. A potent inhibitor of the activated lysate was found to be anti-thrombin II, but specific absorption of anti-thrombin II from plasma reduced only marginally the inhibition of lysate by plasma. The presence of specific antibody to the endotoxin was found not to affect its ability to activate lysate. Inactivation of endotoxin by serum enzymes was biphasic in unheated serum, and most of the activity was destroyed in 3 h at 37 degrees C or in 24 h at 5 degrees C. The relevance of these findings to the objective quantitation of endotoxin activities is discussed.