The ultrastructural study of membrane organization in gram-positive bacteria related to the OSO4 fixation conditions revealed that large, complex mesosomes are observed only when the bacteria are subjected to an initial fixation with 0.1%OSO4 in the culture broth, as in the prefixation step of the Ryter-Kellenberger procedure. Evidence was obtained suggesting that the large mesosomes are produced by this prefization. The kinetic study of the membrane morphological alterations occurring during the prefixation of Bacillus cereus with 0.1%OSO4 in the culture broth showed that the amount of mesosome material increases linearly from zero to a maximum observed at 1.7 min of prefixation and that at about this time a maximum is reached for the number of mesosomes per unity of cell area and for the average individual mesosome area. The large mesosomes observed in gram-positives fixed by the complete Ryter-Kellenberger procedure would be the result of the membrane-damaging action of 0.1%OSO4. Such damaging action was deduced from the observation thay 0.1%OSO4 quickly lyses protoplasts and induces a quick and extensive leakage of intracellular K+ from B. cereus and Streptococcus faecalis. In support of that interpretation is the observation that in bacteria subjected to several membrane-damaging treatments, mesosome-like structures are seen after three different fixation procedures. In bacteria initially fixed with 1% OSO4, 4% OSO4 or 2.5% glutaraldehyde, no large complex mesosomes are observed, small and simple invaginations of the cytoplasmic membrane being present. The size of these minute mesosomes is inversely proportional that causes of fixation. Uranyl acetate was found among the studied fixatives the one to the rate the least damage to bacterial membranes. This fixative satisfactorily preserves protoplasts. In bacteria initially fixed with uranyl acetate no mesosomes were found. The results of the present work throw serious doubts on the existence of mesosomes, both large and small, as real structures of bacterial cells. It is proposed that a continuous cytoplasmic membrane without infoldings (mesosomes) would be the real pattern of membrane organization in gram-positives.