Enterotoxigenic strains of Escherichia coli have been extensively studied over the last ten years, especially in the calf. On the other hand, septicemic strains have received much less attention during the same period, in spite of a lack of information about their pathogenicity in the calf and about the attributes of their virulence. Virulence of septicemic E. coli includes invasion from the portal entry into the blood, multiplication in blood and organs, resistance to phagocytosis and lethal action of serum, and production of lesions by toxins. Most information obtained recently concerns the last phases of microbial pathogenicity, i.e. multiplication and production of lesions. Antigens K are considered to be mainly responsible for the resistance to phagocytosis. Antigens O would play the major role, qualitatively and quantitatively, in the resistance to bactericidal effect of serum. The ability to grow in iron-deficient medium would constitute a decisive advantage. This ability is attributed to proteins of the outer membrane, some of them being coded by plasmids. Endotoxin is the basic toxic principle that accounts for the production of lesions in the host. It is generally admitted that its toxic effect is due to the lipid A portion of LPS. The biochemical composition of lipid A being very homogeneous in Enterobacteriaceae, its pathological effect is considered to be directly related to the concentration in serum, i.e. without specific endotoxic activity exerted by some serotypes. Numerous clinical and experimental arguments suggest the occurrence of an enterotoxemic form of colibacillosis in calves. Enterotoxemic colibacillosis would result from the strictly intestinal multiplication of some strains of E. coli and the systemic diffusion of corresponding endotoxins. Such a syndrome has been well-documented in piglets, namely oedema disease. Oedema disease is attributed to a limited number of serotypes. Experimental reproduction has led to the evidence that two toxic principles are needed to fully produce the disease: the endotoxin and a neurotoxin, called EDP. These results, together with observations on humans and other animal species, raise two major questions: 1) what are the circumstances that favour the passage of endotoxins through intestinal epithelium ? 2) could a specific endotoxic activity be attributed to some serotypes of E. coli?