Oral rabies virus (RV) vaccines are used to immunize a diversity of mammalian carnivores, but no single biological is effective for all major species. Recently, advances in reverse genetics have allowed the design of recombinant RV for consideration as new vaccines. The objective of this experiment was to examine the safety, immunogenicity and efficacy of recombinant RV vaccines administered to captive dogs by the oral route, compared to a commercial vaccinia-rabies glycoprotein (V-RG) recombinant virus vaccine. Animals consisted of naive purpose-bred beagles of both sexes, and were 6 months of age or older. Dogs were randomly assigned to one of six groups, and received either diluent or vaccine (PBS; V-RG; RV SN10-333; RV SPBN-Cyto c; RV SPBNGA; RV SPBNGAGA), with at least six animals per group. On day 0, 1 ml of each vaccine (or PBS) was administered to the oral cavity of each dog, at an approximate concentration of 10(8) to 10(9) TCID50. After vaccination, dogs were observed daily and bled weekly, for 5 weeks, prior to RV challenge. No signs of illness related to vaccination were detected during the observation period. Excluding the controls, RV neutralizing antibodies were detected in the majority of animals within 1-2 weeks of primary vaccination. Thereafter, all dogs were inoculated in the masseter muscle with a street virus of canine origin. All control animals developed rabies, but no vaccinates succumbed, with the exception of a single dog in the V-RG group. Review of these preliminary data demonstrates the non-inferiority of recombinant RV products, as concerns both safety and efficacy, and supports the suggestion that these vaccines may hold promise for future development as oral immunogens for important carnivore species, such as dogs.