ABSTRACT Many molecular techniques have been used to supplement the classification system,which earlier was only based on morphological characterizations. The Amplified Fragment Length Polymorphism (AFLP) marker was successfully used for comparing and identifying locally grown Omani banana cultivars. AFLP technique was used successfully in eighteen Omani banana cultivars from AlBatinah, Al Dhakhilia and Dhofar regions. Eleven AFLP primer combinations were used to develop the fingerprinting of the banana cultivars. Unweighted Pair Group Method with Arithmetic mean (UPGMA) cluster using Jaccards similarity coefficient analysis showed the three distinct banana taxa. Cluster one was banana samples from Al Dakhiliah region which included Bahri, Omani, Maisori Fardh, Sokari and Zanzibar. Second cluster was banana samples from Dhofar region which included Dwarf Spotted Cavendish, Somali, Abubakar Philipino, Maisori Fardh, Milk Banana, Plantain Kenya and Sawara Red. Third cluster was banana samples from Al Batinah region which included Williams, Somali, Malindi, Red Banana, Maisori Fardh and Nagal. Results show an appreciable amount of genetic diversity among the banana cultivars. By using Numerical Taxonomy and Multivariate Analysis System (NTSYS-PC 2.2), a total of 1397 alleles were scored, of which 1322 (94.68%) appeared polymorphic. The primer combination E-ACT/M-CAC showed more percentage of polymorphism alleles about 98.15 % while, E-ACA/ M-CTG