<p>Genomic data are increasingly used in primatology to understand how landscape features, dispersal patterns, and historical processes shape population structure and conservation-relevant diversity. The gelada (<i>Theropithecus gelada</i>), an endemic primate of the Ethiopian highlands, is currently divided into three subspecies; yet little is known about the extent of genomic variation within these groups. This gap is most pronounced for the Central gelada (<i>T. gelada</i> <i>obscurus</i>), the most widespread subspecies, for which genomic data have until now been available from only three individuals from a single site. To broaden this representation, we generated whole-genome sequence data from ten Central gelada individuals sampled across four central Ethiopian sites and two Southern geladas (<i>T.</i> <i>gelada arsi</i>), doubling the number of available genomes for this recently Endangered subspecies. Combining these new data with previously published genomes, we assessed patterns of genomic diversity across the species and tested how major geophysical features of the Ethiopian highlands influence population connectivity. Models incorporating preferred altitudinal ranges for geladas provided a substantially better fit to observed genomic distances than models based only on straight-line geographic distance, indicating that elevational structure strongly constrains movement. Using this framework, we inferred the affinities of unsampled Northern populations, finding stronger support for an east-west than a north-south division between the Northern gelada (<i>T. gelada</i> <i>gelada</i>) and Central gelada. These results highlight the importance of highland topography in shaping gelada population structure and inform future research and conservation planning for Ethiopian primates.</p>