The aim of the dissertation was to undertake phenotypic and genetic characterization and design community-based breeding programs for two indigenous goat populations in Ethiopia, namely Arab and Oromo. The dissertation was based on five manuscripts. Papers I, II, IV and V were based on data generated from production system, morphological characterization and ranking experiments. Paper III was based on whole genome data generated from three indigenous goat populations (Arab, Fellata and Oromo). Genomic data of ten additional goat breeds including four Ethiopian (Abergelle, Gumuz, Keffa, Woiyto-Guji), four African (Kenyan Boran, Morrocan Unknown, Malawian Thyolo, Malian Guera), one Asian (Chinese Tibetan) and one European (French Saanen), obtained from the database, were included for comparison purposes. Reconnaissance survey, focus group discussion and interview with structured questionnaires were used to collect data from 249 households (Paper I). Ten qualitative and nine quantitative traits were recorded/measured from 747 randomly selected goats (Paper II). Genetic diversity metrics within each of three goat populations were assessed and population structure was investigated using principal component analysis (PCA), ADMIXTURE and TREEMIX. Candidate genes underlying strong selection signatures were identified by combining two approaches: pooled hetrozygosity ( Hp) and population differentiation (Fst) (Paper III). Two live animal ranking experiments (own flock and group) were conducted to identify breeding objective traits of Arab and Oromo goat keepers. In the ownflock ranking experiment, a total of 147 households ranked their first, second and third best and the most inferior does in their own flock. In the group-animal ranking experiment, 24 breeding does and bucks were randomly selected and placed into eight groups. Twelve farmers from Arab and Oromo goat keepers, respectively with no knowledge of the experimental animals, ranked the three goats in each group as 1 st , 2 nd and 3 rd , giving reasons for their ranking. After the first round of ranking, the farmers were then provided with the history of each individual goat and asked whether they would reconsider their ranking. This was iterated eight times until a person covered all groups of goats (Paper IV). Four alternative breeding schemes were simulated and evaluated for the Arab and Oromo goat populations (Paper V). The survey results indicated that goats were kept for a variety of purposes. Income generation (0.40), meat (0.23), and savings (0.16) were the highest priorities, but with different index values between the Arab and Oromo goat keepers. Out of the total flock sizes, breeding does constituted the largest (36.7%) followed by kids (23.7%), young does (20.5%), young bucks (8.5%), bucks (7.3%), and castrates (3.4%), even though there was significant (p < 0.01) difference between the two groups of goat keepers. Arab goat keepers owned higher average flock size than Oromo goat keepers. Mating was predominantly uncontrolled mainly due to communal grazing. On average, first kidding occured at the age of 1.2 years, kidding interval at 7.5 months, reproductive life was for about 7.6 years, and produced 10.7 kids per lifetime despite variation between Arab and Oromo goat keepers (Paper I). The qualitative and quantitative data analysis revealed significant (p < 0.01) differences between Arab and Oromo goats. White coat color was predominantly observed in Arab goats (33.72%) while brown (deep and light) coat color was the most frequent in Oromo goats (27.81%). Goat population type and age had significant effects on body weight and linear body measurements. When goats of the same age class were compared, Oromo goats were heavier and bigger in size. Positive and highly significant correlations were obtained between body weight and most of the body measurements in both goat populations. The highest correlation was between chest girth and body weight for Arab (r = 0.95) and Oromo (r = 0.92) (Paper II). The genomic data analysis generated 35,161,094 autosomal biallelic single nucleotide polymorphisms (SNPs) and 3,737,445 insertions and deletions (InDels) from the 13 global goat populations (78 goat genomes). The highest number of SNPs and InDels were detected in Unknown goat while the lowest number of SNPs and InDels were identified in Saanen and Gumuz goats, respectively. Keffa goat demonstrated the lowest level of genetic diversity whereas Woyito-Guji showed the highest. Our results show an inverse relationship between the length and number of runs of homozygosity (ROH). The number of shorter ROHs (ROH 100–150Kb ) category predominated; accounting for 58.18% of the total number of ROHs. The longest ROH category (ROH >400Kb ) was the rarest, wherein the Gumuz population displayed the highest number of ROH, suggesting recent inbreeding in this population. Principal component analysis (PCA), ADMIXTURE and TREEMIX tools revealed five distinct genetic clusters in the global, three in the African and two in the Ethiopian goat populations. For the selection signature analysis, we used comparative genomics to explore differences between populations that are exposed to different selection pressures (natural versus artificial) and from different geographic zones (tropical semi-arid versus temperate cold) using pooled hetrozygosity (Hp) and population differentiation (Fst) between the following population pairs: Gumuz versus Saanen and Fellata versus Saanen. By combining the two approaches, candidate genes underlying strong selection signatures including litter size, adaptation to arid environment, coat coloration and body weight were identified (Paper III). The live animal ranking experiments indicated that traits like body size, twining ability, kidding interval and mothering ability were the topmost preferred attributes to select breeding does; whereas body size, libido and growth were the traits used to select breeding bucks in the study area (Paper IV). The four alternative breeding schemes simulated for Arab and Oromo goats resulted in reasonable annual genetic gain (PAGG), especially for six months weight (6 mw). Due to its economic return, Scheme 2 is recommended over other schemes (Paper V). In conclusion, future works should focus on the implementation of CBBP in the study area using Scheme 2. However, care should be taken to ensure the participation of goat keepers from inception through to implementation and include few priority breeding objective traits.