We evaluate how the environmental complexity of La Malinche volcano influences patterns of genomic variation in Peromyscus melanotis, across two mountain slopes and global and regional geographic scales.Using reduced representation genomic sequencing we estimated population genetic diversity, subdivisions and migration rates. Remote sensing data and drone image processing were used to characterize landscape variables. We evaluated their effect on connectivity, based on a landscape analyses framework, using resistance surfaces, circuit theory and omnidirectional connectivity. Our findings showed how the forest environmental complexity across different geographic scales drives dispersal, genomic structure and connectivity patterns in this rodent, where dirt roads and disturbed areas limit its connectivity, while exhibiting higher connectivity at the highest elevations where the forest is less disturbed. Notably, that our 3D variable of tree height was significant demonstrates the utility in incorporating 3D vegetation structure variables into landscape genetic analyses.