In view to furnish data to elucita the controversial problem of the nature of Cytamoeba bacterifera, which it is often found in erythrocytes of the frog Leptodactylus ocellatus, we have done some cytochemical tests. We displayd in its structure the contents of RNA and polysaccharides that do not undergo digestion by ptyalin. With the Feulgen's method we mainly obtained negative results, but one parasite showed a faintly diffused positivity and a few doubtful reaction were also observed. When we employed the Methyl Green-Pyronin, even after the ribonuclease treatment, and the Toluidin Blue, likewise subsequent to the activity of this enzime, we did not evidence the DNA. Taking into account that the components of C. bacterifere are usually very small and sometimes undiscernible, it is probable that their contents of DNA, perhaps existent, would be difficultly demonstrated under the light microscope, and also that it would be easily dissimulated by other substances. We do not believe in a viral nature of Cytamoeba mainly regarding some aspects of its structure (figs. 8, 15, 17 and 18) and the absence of damage on the nucleous of the infected cells. Although we did not prove the presence of DNA, we think that it is possible that C. bacterifera is an intracytoplasmic clump of modified organisms wich sizes are just within the nearness of th limit of visibility with optical microscopes, related with the bacteria, such as are, by exemple, the Chlamydiae and Rickettsiae. The development of Cytamoeba was followed in a frog maintained int he laboratory during three months and weekly examined; we observed progressive decrease of the parasitismo. Its structural types and sizes were not related with the phasis of the infection. We did not get transmission by intraperitonela inoculation of C. bacterifera from L. ocellatus to Bufo crucifer.