Tef (Eragrostis tef (Zucc.) Trotter) is a gluten free cereal crop mainly grown in Ethiopia for food, feed and cash. Knowledge on the nature and magnitude of variation existing in breeding materials and interrelationships between traits of great importance to effective breeding. This study was designed to assess the genetic variability using morphological traits and to investigate trait association among 64 tef genotypes. The genotypes were laid out in 8x8 simple lattice design at Adet agricultural research centre in rainy season of 2019/20. Data were collected on grain yield and 11 yield-related traits and 3 qualitative traits. The analysis of variance revealed that highly significant (P≤ 0.01) and significant (P≤ 0. 05) differences among the tested genotypes for all recorded traits considered, identify existence of genetic variability that can be exploited in future tef breeding program. A wide range of variation was observed in days to heading, days to maturity, plant height, panicle length, culm length, number of effective tillers, unproductive tillers, number of total tillers, biomass yield, grain yield, harvest index and lodging index. Moderate to high phenotypic and genotypic coefficients of variation were recorded for days to maturity, culm length, number of effective tillers, number of total tiller, grain yield, harvest and lodging index. Moderate to high heritability coupled with genetic advance percent of mean was observed for days to maturity, grain yield, lodging and harvest index and tillers (number of effective tillers, number of total tillers). This result implied, these traits could be improved by direct selection. Grain yield was positively and significantly correlated with number of effective tillers, number of total tillers, plant height, biomass yield and harvest index both at the phenotypic and genotypic levels. The path coefficient analyses revealed that the dominant component traits determining grain yield were plant height, number of effective tillers, number of total tillers, biomass yield and harvest index. These traits could be used as indirect selection criteria to improve tef grain yield. Cluster analysis showed genotypes were grouped to four clusters of the 64 tef genotypes. The highest inter cluster distance was recorded between clusters III and IV; as a result, tef genotype from these clusters could be used as parents for future crossing programs. Principal components analysis revealed that the first four principal component axes accounted for 75.12% of the total variation among genotypes, indicating that the traits considered were appropriate to detect variation among tested genotypes Overall, the present study confirmed the presence of considerable variability among genotypes; and therefore, promising genotypes should be considered be future tef improvement programs, however the study needs to test in different location in more than one year. Keywords: Cluster analysis, Correlation, Genetic Advance, Heritability, Path Analysis, Principal Component.