Phenotypic and isozyme diversity of 32 landrace populations of tetraploid wheats originatingrnfrom two regions (Bale and Wello) of Ethiopia was analysed in this study. A total of 2559rnindividual plants (45-110 plants pel' population) representing each landrace were consideredrnfor the phenotypic diversity investigation. Investigations were also made on the speciesrnrichness (biological diversity) of the popUlations whereby individuals in each landrace werernclassified into their own taxonomic classes and analysis of the species diversity was made forrneach population, altitudinal class and region of collection. Seven heritable qualitative traitsrn(seed colDl; glume color, awn color, glume hairiness, beak awn, spike density, and awnrnlength) were studied. The frequencies of each phenotypic class were used to compute thernShannoncWeaver index (H') to estimate and analyse the diversity at different levels. Forrnisozyme analysis, three enzyme systems, Esterase (EST), Aspartate amino trans ph erasern(AAT), alld Leucine amino peptidase (LAP) were used to study the genetic variatioll in 14rn- .rnlandrace populations selected based on the clustering of the populations from the phenotypicrnfrequencies. Each population was represented by twenty individuals.rnThe analysis of species diversity indicated that seven species of wheats from two ploidy levels,rntetraploid and hexaploicl, are planted in mixtures at vG/ying proportions, Triticum turgidumrnssp. durum was the most dominant species accountillg for about 67.32 % of the totalrnpopulation. A maximum of four species was recorded per each population. Analysis ofrnphenotypic frequencies and diversity indices were done for tetraploids lumped together andrnfor durum alolle. Considering tetraploid species, seed color alld beak awn contributed thernhighest diversity index (H' = 0.92 for both). Awn color contributed the lowest diversity indexrn(H' = 0.59). When durum treated alone, glume color followed by beak awn and seed colorrncontributed the highest diversity index (H' = 0.90, 0.84, alld 0.84, respectively), whereasrnglume hairiness had the 10IVest diversity index (H' = 0.48).rnTraits sllch as awn color, awn length, glume color and seed color showed statisticallyrnsignificant differences (p < 0.05) between regions while significant differences (p < 0.05)rnwere observed for awn length, beak awn, glume hairiness and seed color across gradients of altitudillal classes in tetraploid wheats. Sigllificallt differences were also observed betweellrnaltitudes and regions for traits in dU1'II1ll treated alone. Variations were, however, largely duernto the differences in the level of the different characters within populations. On regionalrnbasis, higher lIIean diversity index was observed for Wello than Bale. Awn color wasrnmonolllOlphic and jixed for color types in collections originating fi'om Bale whereas a varietyrnof fi'equencies were observed in Wello. Considering tetraploids, the overall diversity indexrnwas cOlllputed to be H' = 0.81 Â± 0.08. The overall region value of H' for durum wheat is 0.63rnÂ± 0.08 indicating a decline in the alllount of diversity in durulII wheat as compared with thernindex for tetraploid wheats.rnThe isozyme assay resulted into a total of 5 isozyme loci; 2 of the distinct polymO/phicrnregions of enzyme activity (banding paUe/'ll) detected for EST, 2 for AAT, and one fO/; LAP.rnThe isozyme data also indicated presence of allelic polymO/p/zism was detected at all loci.rnThe mean number of alleles per locus ranged fi'om 1.4 to 2.2. Considerillg the geneticrnvariability at the 5 loci, the percentage polymO/phic loci varied fi'om 40% to 80%. A pairwiserncomparisoll of populations over jive isozyme loci indicates that the ullbiased geneticrnidentity rangedfi'om 0.904 to 1.00. Whereas the gelletic distallce rallgedfi'om 0.00 to 0.053,rnindicatillg that there is small amount of variatioll amollg populatiOlls. Populations wererngrouped/clustered illto six clusters. The landraces ji'om Bale and Wello grouped differentlyrnsimilar to that of clustering that was based on phenotypic frequencies.