Anopheles gambiae, Anopheles coluzzii, and Anopheles arabiensis are three of the most widespread vectors of malaria parasites, with geographical ranges stretching across wide swaths of Africa. Understanding the population structure of these closely related species, including the extent to which populations are connected by gene flow, is essential for understanding how vector control implemented in one location might indirectly affect vector populations in other locations. Here, we assessed the population structure of each species based on a combined data set of publicly available and newly processed whole-genome sequences. The data set included single nucleotide polymorphisms from whole genomes of 2,410 individual mosquitoes sampled from 128 locations across 19 African countries. We found that A. gambiae sampled from several countries in West and Central Africa showed low genetic differentiation from each other according to principal components analysis (PCA) and ADMIXTURE modeling. Using Estimated Effective Migration Surfaces (EEMS), we showed that this low genetic differentiation indicates high effective migration rates for A. gambiae across this region. Outside of this region, we found eight groups of sampling locations from Central, East, and Southern Africa for which A. gambiae showed higher genetic differentiation, and lower effective migration rates, between each other and the West/Central Africa group. These results indicate that the barriers to and corridors for migration between populations of A. gambiae differ across the geographical range of this malaria vector species. Using the same methods, we found higher genetic differentiation and lower migration rates between populations of A. coluzzii in West and Central Africa than for A. gambiae in the same region. In contrast, we found lower genetic differentiation and higher migration rates between populations of A. arabiensis in Tanzania, compared to A. gambiae in the same region. These differences between A. gambiae, A. coluzzii, and A. arabiensis indicate that migration barriers and corridors may vary, even between very closely related species. Overall, our results demonstrate that migration rates vary both within and between species of Anopheles mosquitoes, presumably based on species-specific responses to the ecological or environmental conditions that may impede or facilitate migration, and the geographical patterns of these conditions across the landscape. Together with previous findings, this study provides robust evidence that migration rates between populations of malaria vectors depend on the ecological context, which should be considered when planning surveillance of vector populations, monitoring for insecticide resistance, and evaluating interventions.