This study was conducted to evaluate the efficiency of SRAP, TRAP and SSR primers in detecting genetic variation among 17 C. arabica genotypes collected from the different valleys of Yafea City, Yemen, and estimate genetic similarity coefficients among these genotypes and classify them according to genetic relationships. Results of the present study demonstrated the presence of genetic variation among coffee genotypes within and between valleys. Cluster analysis using SRAP, TRAP, SSR and combined data, grouped the 17 coffee genotypes into three main clusters with Jaccard's similarity coefficient ranging from 0.81 to 0.98, from 0.90 to 0.99, from 0.69 to 0.90 and from 0.84 to 0.96, respectively. The combined analysis produced similar dendrogram to that produced using SRAP marker analysis. Sixteen SSR primers were used to investigate the discriminative power of each SSR primer by calculating polymorphic information contents. A total of 37 alleles were amplified among 17 coffee genotypes using sixteen SSR primers. The number of amplified alleles per primer varied from one allele as for AJ308738 to seven alleles for AJ250256, with an average value of 2.31 alleles. The polymorphic information content per primer ranged from 0.000 to 0.713 with an average of 0.319. Observed heterozygosity was calculated for each primer and ranged from 0.00 to 1.00 with an average of 0.43.