Ethiopia is the center of origin and diversity for Arabica coffee with high morphological diversity between accessions as compared to commercial cultivars. Coffee germplasm collection and molecular characterization are crucial steps towards its conservation, breeding, and development of superior genotypes for various end uses. Hence, this study was initiated with the objective of studying the genetic diversity of Coffea arabica accessions collected from different regions of Ethiopia, using SSR markers. A total of 20 SSR markers were used to genotype 86 accessions and produced a total of 112 alleles, with an average of 5.6 alleles per locus. All the loci across the entire populations were found to be highly polymorphic with a mean of 0.6 PIC value. Average observed heterozygosity and allelic richness across all populations ranged from 0.22 to 0.27 and 3.52–4.26, with a mean of 2.43 and 3.97, respectively. AMOVA showed high variation within population based on geographical origin. The smaller Fst (0.037) observed indicates the presence of lower population genetic differentiation as a result of higher gene flow (Nm = 2.45) across populations and the lowest mean genetic distance (0.21) observed between populations. The UPGMA, PCoA, and structure analysis poorly grouped the individuals into distinct clusters indicating the presence of population admixture. The observed higher genetic variability in all populations indicates that the country has huge coffee genetic diversity which can be used for future coffee improvement. Our results revealed an unexploited highly diverse genetic resource particularly from Omo, Ilubabor, and Benchi Maji that should be considered in future coffee breeding program and germplasm conservation.