We compared three vaccination strategies in three rural communities in Mexico to determine the factors associated with the success of vaccination programs in areas where canine rabies is poorly controlled. In town A, intensive publicity and community participation were used; owners were instructed to bring their dogs to temporary centralized clinics for vaccination. In town B, only brief precampaign publicity was used, followed by vaccination at a centralized site. Minimal publicity was also used in town C, but the vaccination campaign was conducted house to house. A total of 5,426 residents and 1,597 dogs were counted in the three towns (mean human:dog ratio 3.4:1). In Town A, 70.1% (472 of 673) of the dogs were vaccinated; the campaign required 40 person-minutes per dog. Significantly greater proportions were vaccinated in town B (262 of 318 [82.4%]; P less than 0.001) and town C (483 of 561 [86.1%]; P less than 0.00001); each of these latter campaigns required 10 person-minutes per dog. The following factors were positively associated (by multivariate analyses) with vaccination of individual dogs: non-intensive publicity, house-to-house vaccination, dogs owned by a single member of the household, and dogs acquired greater than 15 days after birth. Intensive publicity did not increase the overall success of the vaccination program; the efficiency of centralized versus and house-to-house vaccination was comparable.