Sorghum (Sorghum bicolor (L.) Moench) is one of the most important multipurpose crops grownrnglobally. It can be used as food, feed, fiber and biofuel feedstock. In Ethiopia, sorghum is amongrnthe most important cereal crops accounting for 18 % of annual production and 16 % of landrncover allocated to the entire grain crops including cereals, pulses and oilseeds. Despite thisrnimportance and huge potential as bioenergy feedstock, detailed study on characterization,rnidentification and mapping of novel genes that code for sugar content related and droughtrntolerance traits using Ethiopian sorghum was lacking. Therefore, this study is aimed atrngermplasm collection, characterization, and identification and mapping of novel QTLs using bothrngrain and sweet type of sorghum germplasm of Ethiopia. Both morphological and molecularrnevaluation of Ethiopian sweet sorghum germplasm was undertaken. Genotyping of 175 Ethiopianrnsweet sorghum genotypes alongside 27 improved accessions from eastern and southern Africarnwas undertaken. Two independent experiments were also carried out to identify and map QTLsrnassociated to Brix and stay-green related traits using F2:3 segregating mapping populationsrnderived from a cross between grain sorghum with stay-green feature (Sorcoll 163/07) and sweetrnsorghum (Gambella). A genetic map was constructed using 192 F2 populations genotyped with 76rnSSR markers. Research was carried out to screen and compile the most informative SSR markersrnacross some accessions. A total of 304 markers were used to identify the most polymorphic SSRrnmarkers across eleven farmers preferred sorghum genotypes. Combined analysis of variance forrnBrix and other morphological characters to evaluate 181 sweet sorghum accessions collectedrnfrom the major producing regions of Ethiopia showed that there is a significant effect of thernenvironment on all tested traits. Mean separation analysis revealed that collections from northernrnEthiopia were found to be superior in terms of Brix degree. Collections from the rest of collectionrnregions showed relatively low Brix mean value but characterized by higher biomass. Broad sensernheritability estimate showed most characters are highly heritable except grain yield. Pearsonrncorrelation coefficients (r) of the seven traits presented Brix degree was negatively correlated tornmost of the traits. Cluster analysis based on the Brix and other morphological characters groupedrnthe accessions into five clusters. The constructed dendrogram based on mean of collection zonesrnfor the tested traits also clearly put adjacent regions or zones together. All the tested markersrndetected 159 alleles and a high degree of polymorphism information content (PIC) averagingrn0.69. A comparison between Ethiopian and improved accessions revealed higher allele numbersrn(124) in Ethiopian than improved accessions (92 alleles). More than half (65 out of 124) of thernalleles observed in the Ethiopian accessions were rare (