Trials were performed using the V-RG(B+C) and SAG1(A+B) vaccines. No clinical signs or lesions were recorded in any of the inoculated foxes during, 1, 6, 12 and 18 months after vaccination. The absence of pathogenicity was observed whatever the dose inoculated or route of administration. The absence of pathogenicity of V-RG administered by the oral route also has been demonstrated in laboratory and domestic animals. Moreover, V-RG innocuity was tested in several wild animal species. Several species of predatory or scavenger birds, carnivores, rodents and hoofed mammals. Postmortem examination did not result in the detection of lesions. During a minimum of 28 days postvaccination, neither clinical signs were observed in any of vaccinated animals. Sero conversion has never been observed in unvaccinated controls held in close contact with V-RG vaccinated ones (red fox, dog, cat, cattle, ferret, badger and wild boar). Several laboratory techniques failed to detect any virus in the saliva, salivary glands, the blood and internal organs of freshly inoculated foxes. The method of polymerase chain reaction could detect the V-RG virus at very low levels in tonsils, buccal mucosa and soft palate only. Finally, the genetic stability of this recombinant virus could be attested after serial passages in vitro (cell lines) as well as in vivo (red fox, laboratory mouse). Stability of the SAG1 was assessed as absolute in cell culture. But attenuation results from a single mutation and is then subjected to reversion (as it was seen in wild vole after one intracerebral passage). Most likely the revertant will not induce rabies, because its multiplication will be restricted by the development of the immune response. It will then remain trapped in the vaccine and will disappear with him. An experimental study has shown that early or late death phenomenon, as consequences of interactions between oral vaccination with V-RG and rabies infection, is likely to impede the creation of asymptom atic carriers of rabies virus.