Summary Whether spore-forming or non-spore forming, some anaerobes are important pathogens to consider by reason of their presence in the environment and in the human natural cavities. We will discuss new data in clostridia which are characterized by their ubiquity, their ability to survive as spores under adverse conditions and their capacity to produce potent toxins. New toxins have been recently described, i.e. an hemolysin in C. septicm and a larvicidal toxin, of potential interest as an anti-vectorial agent, in C. bifermentans . A C. butyricum strain and a C. barati strain producing botulinal toxins of type E and F, respectively have been isolated in infant botulism; also a C. botulinum strain producing two toxins, A and B has been isolated in a botulism outbreak. Taken together, these observations raise the question of the genetic determinism underlying the occurrence of toxins amongst previously non toxinogenic clostridial species. Infant botulism is an old disease, but a recently described clinical entity. The precise mechanisms underlying this toxi-infection are actually unknown, and probable involve a temporary immunity deficiency in some newbornsk. In C. difficile , several points have been partially solved. There is now a consensus concerning the existence of two toxins of high molecular weight, A and B that cross react against the present available antisera and thus should share common epitopes. Cloning studies suggest that gene(s) coding for both toxins should be very close, if not identical. Both toxins apparently have a similar mode of action (cytotoxic activity), the toxin A having in addition an enterotoxin action. The receptors for each toxin should be different, and should also differ in man and animal. Several questions are presently still unsolved, particularly; (i) the question in C. difficile of additional virulence determinants; (ii) the lack of virulence of C. difficile in newborns: the lack of specific toxin(s) receptors in the newborn has been suggested, but has yet to be confirmed; ( iii ) the role of antibiotic administration in the onset of the disease: if, in some cases, the administered antibiotic may directly select a C. difficile strain resistant to this antibiotic, in other cases, the antibiotic would exert an indirect action by inhibiting bacterial flora involved in resistance to colonization by C. difficile . The role of Clostridia , and particularly of C. butyricum in neonatal enterocolitis (NNE) is presently a matter of controversy. Epidemiological and experimental evidence have raised the question of the potential role of C. butyricum in NNE, but this species was also found in control infants and no specific toxicity determinant of C. butyricum has been shown in experiments with mice or culture cells, except a thermostable factor of low molecular weight identified as butyric acid. An experimental axenic chicken/quail model was then developed. Whether it is possible to extrapolate the chicken/quail model to human is a mute point. Virulence factors of non-spore forming anaerobes are discussed in most of the communications of the present symposium, and they are dealing with adherence, tissu damaging products, and host-protective properties. We wish just to emphasize the recent characterization of enterotoxigenic Bacteroides fragilis strains causing severe diarrhea in man and livestock, and experimentally capable of producing an enteric disease in infant rabbits. Transferable antibiotic resistance in anaerobes, either in Clostridium or in Bacteroides , are now well documented and such resistances may be considered as virulence cofactors. The genetic determinants of the antibiotic resistances are chromosomally-determined, plasmid-borne or localized on transposons. A relatively great variety of mechanisms of horizontal transfer of these genetic determinants have been identified in those anaerobes: self-mobilizing plasmids in C. perfringens and Bacteroides , conjugal Tc R and Tc R Cc R elements, conjugative transposon-like elements, and a conjugal mobilizing transposon in Bacteroides . In other groups, i.e. C. difficile and C. innocuum chromosomally-determined antibiotic resistance are transferable through a mechanism presently uncharacterized. Recent epidemiological data suggest the spreading of antibiotic resistance in other groups, i.e. the anaerobic cocci.