ObjectiveTo evaluate the efficacy of a synthetic sex-aggregation pheromone of the sand fly vector Lu. longipalpis, co-located with residual insecticide, to reduce the infection incidence of Leishmania infantum in the canine reservoir.MethodsA stratified cluster randomised trial was designed to detect a 50% reduction in canine incident infection after 24 months in 42 recruited clusters, randomly assigned to one of three intervention arms (14 cluster each): synthetic pheromone + insecticide, insecticide-impregnated dog collars, or placebo control. Infection incidence was measured by seroconversion to anti-Leishmania serum antibody, Leishmania parasite detection and canine tissue parasite loads. Changes in relative Lu. longipalpis abundance within households were measured by setting three CDC light traps per household.ResultsA total 1,454 seronegative dogs were followed-up for a median 15.2 (95% C.I.s: 14.6, 16.2) months per cluster. The pheromone + insecticide intervention provided 13% (95% C.I. 0%, 44.0%) protection against anti-Leishmania antibody seroconversion, 52% (95% C.I. 6.2%, 74·9%) against parasite infection, reduced tissue parasite loads by 53% (95% C.I. 5.4%, 76.7%), and reduced household female sand fly abundance by 49% (95% C.I. 8.2%, 71.3%). Variation in the efficacy against seroconversion varied between trial strata. Equivalent protection attributed to the impregnated-collars were 36% (95% C.I. 14.4%, 51.8%), 23% (95% C.I. 0%, 57·5%), 48% (95% C.I. 0%, 73.4%) and 43% (95% C.I. 0%, 67.9%), respectively. Comparison of the two interventions showed no statistically consistent differences in their efficacies; however, the errors were broad for all outcomes. Reductions in sand fly numbers were predominant where insecticide was located (chicken and dog sleeping sites), with no evidence of insecticide-induced repellence onto humans or dogs.ConclusionThe synthetic pheromone co-located with insecticide provides protection particularly against canine L. infantum parasite transmission and sand fly vector abundance. The effect estimates are not dissimilar to those of the insecticide-impregnated collars, which are documented to reduce canine infection incidence, human infection and clinical VL disease incidence, in different global regions. The trialled novel lure-and-kill approach is a low-cost potential vector control tool against ZVL in the Americas.