Most investigations of geographical differences within microbial species are limited to focusing on a single species. Here, we investigate the global differences for multiple bacterial species by using a dataset of 757 metagenomics sewage samples from 101 different countries worldwide. The within-species variations were identified by performing unsupervised genome reconstructions that reduce database and mapping biases, and the analyses were further expanded by using gene focused approaches. Applying these methods, we recovered 3,353 near complete (NC) metagenome assembled genomes (MAGs) encompassing 1,439 different MAG species and found that within-species genomic variation was often coherent with regional separation. Additionally, we found that the variation of organelle genes correlated less with geography compared to metabolic and membrane associated genes, suggesting that the global differences of these species are caused by regional environmental selection rather than limitations on dissemination. From the combination of the large and globally distributed dataset with the in-depth analysis methods, we present the most comprehensive investigation of global within-species phylogeny from metagenomics data to date.