- East Africa is considered to be a secondary centre of origin of sweet potato and it is suspected that the wide morphological variation observed indicates wide genetic diversity in the region. To conserve and utilize the germplasm, it is important that proper assessment of the diversity of the East African sweet potato germplasm be made.Identification by molecular technologies is more commonly used over morphological characters since the latter can be influenced by environmental factors. In this study, we used molecular and morphological markers to study the genetic diversity of the germplasm in the region. Collections of cultivars were made from selected locations of Uganda, Kenya and Tanzania and subsequently established in pots in a screen house at Makerere University. A total of 266 cultivars were collected. After 3 weeks, the cultivars were screened for morphological characters using the CIP Research Guide. Cluster analysis was done using UPGMA in Treecon (Version 1.3). Based on morphological grouping, 57 cultivars, which were morphologically diverse, were randomly selected for DNA extraction and further analysis was done. Cluster analysis revealed only two major groupings (A & B) of sweet potatoes with very low bootstrap support of 0-54 %. The key distinguishing morphological markers were triangular leaf outline and a cordate shaped leaf outline for group A & B respectively. In addition, there were no geographical distinct morphological types identified. No population structure was detected. However, within each country, a high variation was observed (97.65%), suggesting that a wide range of cultivars is being grown in each country. Microsatellite (SSR) reactions were performed using four SSR primer combinations. The polymerase chain reaction (PCR) products were resolved using a high resolution metaphor agarose gel electrophoresis. Genetic distance data matrices were subjected to Unweighted pair-group method of arithmetic averages (UPGMA) clustering using TREECON phylogenetic program Version 1.3 b. Two major sub-clusters were found by UPGMA at a bootstrap value of 54 %. Low bootstrap values (0-55 %) indicate absence of clusters and close genetic relationships among the cultivars. The majority of cultivars were in the range of 0.1-0.3 Nei's genetic distance from each other, which also shows close genetic relatedness. The clustering of sweet potato cultivars based on SSR markers showed that cultivars from Kenya, Uganda and Tanzania were grouping in group A. In sub-cluster B the cultivars were from Uganda and they seemed to form a unique group. However the Tanzanian cultivars seem to cluster closely together in various sub-clusters. Analysis of Molecular Variance (AMOVA) indicated that there is statistically measurable divergence between the sweet potato of Uganda-Kenya and the other East-African country, Tanzania with detectable difference between the cultivars of the three sources. The largest source of diversity comes from within-population variation, which accounts for 88.91 % of the total variance. The data from AMOVA analysis also indicated an F st value >0.05 which seems to suggest great genetic differentiation amongst the cultivars in the East African region and hence presence of a population structure. The gene flow values > 1 shows that there is high genetic drift amongst the cultivars in this region. In this study, the morphological analysis of sweet potato landraces indicated that there was not much variation in the East African sweet potato. However the investigation at genome level using PCR-based SSR markers was able to identify significant variation amongst the landraces and existence of a population structure. The major results in this study indicate that SSR markers are appropriate for the genotyping and revealing genetic relationship of East African sweetpotato cultivars. In addition, morphological characterisation should be complemented with DNA –based characterisation using SSR markers to reveal genetic diversity of East African sweet potato cultivars.