Over the last decade, the release of Wolbachia-infected Aedes aegypti into the natural habitat of this mosquito species has become the most sustainable and long-lasting technique to prevent and control vector-borne diseases, such as dengue, zika, or chikungunya. However, the limited resources to generate such mosquitoes and their effective distribution in large areas dominated by the Aedes aegypti vector represent a challenge for policymakers. Here, we introduce a mathematical framework for the spread of dengue in which competition between wild and Wolbachia-infected mosquitoes, the cross-contagion patterns between humans and vectors, the heterogeneous distribution of the human population in different areas, and the mobility flows between them are combined. Our framework allows us to identify the most effective areas for the release of Wolbachia-infected mosquitoes to achieve a large decrease in the global dengue prevalence.