A field trial to evaluate the efficacy of an oral vaccinia-rabies glycoprotein recombinant virus vaccine in controlling epidemic raccoon (Procyon lotor) rabies was conducted by distributing 180,816 doses (10(8.2)TCID50/ml) of vaccine in wax ampules within fish-meal polymer baits at a rate of 64 doses/km2/treatment throughout a 552 km2 area, forming an 18 km wide band across the northern Cape May Peninsula of New Jersey (USA). Vaccination treatments were conducted in the spring and fall between May 1992 and October 1994 from a helicopter along ecotones and from motor vehicles along roads. Vaccine-laden baits were removed by animals from tracking stations within 3 wk and 61% of the identifiable tracks were those of raccoons. Tetracycline incorporated in the baits as a biomarker was detected in 155 (73%) of the vaccination area raccoons following the fall 1993 and spring 1994 vaccinations. Eleven (61%) of the raccoons sampled in the same time period seroconverted (> or = 0.5 IU) in response to rabies virus glycoprotein. A raccoon diagnosed with rabies from the northern border of the vaccination area on 30 April 1993 provided the first evidence that the barrier was being challenged by the rabies epidemic. The prevalence of rabies in raccoons from the vaccination area for the first year (10%, n = 96) and second year (8%, n = 61) of challenge was reduced more than six-fold by vaccination compared to unvaccinated raccoons from northern adjacent surveillance areas during the corresponding first (65%, n = 189) and second years (53%, n = 43). Vaccination also effectively reduced by three-fold the rate at which the epidemic moved through the raccoon population (15 km/yr). The breach of the vaccination area resulted in a resumption of the high rate (43 km/yr) of epidemic movement and a significant nine-fold increase in rabies prevalence (77%, n = 47). The maximum linear movement (12.9 km) among five ear-tagged rabid raccoons in the study area was significantly greater than that of 19 normal radio-collared raccoons (2.58 km) in the area. These large movements of rabid raccoons, together with relocation of nuisance raccoons, spillover of raccoon rabies in skunks (Mephitis mephitis) and other species, insufficient funding and a decision to discontinue the program in 1994 (which could have resulted in insufficient population immunity among raccoons in the vaccination area) may have contributed to the eventual breach of the barrier.