Toddia França, 1912 under the light microscope occurs as inclusion corpuscles in the cytoplasm of erythrocytes of cold-blooded vertebrates sometimes accompanied by crystalloid bodies. Its position among the protozoans or the viruses has been discussed by some authors, but remained unclear. To elucidate this problem we studied Toddia from a Brazilian frog (Leptodactylus ocellatus) by electron microscopy. In the cytoplasm of the infected cells we found no protozoan, but rather virus-like particles often hexagonal in outline, averaging 195 nm excluding their two involving membranes, and presenting a central area of variable electron density. Particles at different stages of development were generally found around or on area lighter density than the cytoplasm. which resembled a virus synthesis site. At high magnification, the nuclear or cytoplasmic crystals allied to Toddia resembled the crystalline lattice of the inclusion bodies associated with the polyhedrosis viruses and poxviruses from insects, of the capsules of granulosis viruses and of other protein crystals in ultrathin sections. Cytochemical tests in Toddia corpuscles displayed exclusively the presence of deoxyribonucleic acid. These findings indicate that Toddia is not a protozoan and demonstrate that it is in all probability a viral inclusion corpuscle. Taking into account the nucleic acid type found in its structure (DNA) and the hexagonal shape usually shown in ultrathin sections by its component particles, which have a cytoplasmic site of synthesis and assembly, we tentatively relate Toddia with the so-called "Icosahedral Cytoplasmic Deoxyriboviruses". We believe that the present paper gives the first report of virus-like particles in L. ocellatus.