Mathematical models are used to compare the effectiveness of various untested, unused, and undeveloped methods for controlling canine and human zoonotic visceral leishmaniasis (ZVL), including insecticides, vaccines, killing serologically positive and sick dogs, and drugs. For given percentage changes in control parameters, insecticides are the most effective control method. Where transmission occurs peridomestically and vectors are accessible to treatment, as in parts of tropical America, insecticides are expected to reduce the incidence of human ZVL even more effectively than they reduce the prevalence of canine leishmaniasis, a result that should encourage properly designed vector control trials. The second best strategy is to reduce susceptibility to leishmaniasis by vaccinating people or dogs, or by eliminating childhood malnutrition where it is common. Both killing vectors and reducing susceptibility (by whatever means) are more effective than killing dogs or treating them with drugs. In Europe, where vector control is less likely to be successful and canine leishmaniasis is a major veterinary problem, a dog vaccine is highly desirable. Better drugs for dogs will help case management but, with regard to bringing down the incidence in the dog population, immunization is the ultimate goal.