A vaccination program for the control of terrestrial rabies in the province of Ontario, Canada, began in 1989. During the period between 1989 and 2004, over 13 million baits containing the live, attenuated rabies virus ERA-BHK21 were distributed across the province, with the aim of immunizing foxes by the oral route. Animals recovered from bait distribution areas were assayed by fluorescent antibody test for rabies virus infection. Immunoreactivity with a panel of monoclonal antibodies that discriminate between ERA and rabies virus variants known to circulate in Ontario, and molecular genetic analyses were used to identify animals infected with ERA. Nine cases of ERA variant rabies were identified over the 16-yr period of study; these did not appear to be stratified by species, year of discovery, or location of capture. The ERA-positive animals were found across the province in eight counties, all of which had been baited in the year of case discovery. The nine ERA-positive cases included four red foxes (Vulpes vulpes), two raccoons (Procyon lotor), two striped skunks (Mephitis mephitis), and one bovine calf (Bos taurus). Molecular phylogenetic analyses of the partial N gene sequences generated from these isolates indicated that these nine cases were due to infection with the ERA variant. The glycoprotein sequences were predicted from G gene sequencing of all nine field isolates and two laboratory stock ERA viruses. This revealed some heterogeneity at residue 120 (either arginine or histidine) in both field and laboratory stocks as well as a few other mutations in field isolates. The significance of this heterogeneity remains unclear. Our data demonstrate that the ERA vaccine distributed in Ontario carried residual pathogenicity; however, there does not appear to be any evidence of ERA establishment in wildlife populations over the 16-yr period. These results are consistent with previous reports of the rare detection of ERA vaccine–induced rabies and with laboratory studies of ERA pathogenicity.