Barley breeders need to achieve the aim of developing genotypes that will maximize yield under unfavorable conditions which could be realized through selection for broad and specific adaptation by using different breeding methods to develop superior cultivars. This investigation was initiated to assess the effectiveness of four selection methods, i.e. bulk selection method (BM), modified bulk selection method (MBM), pedigree selection method (PM)and single seed descent selection method (SSDM) applied in the early segregating generations of 10 diverse six-rowed barley crosses to improve yield and its components. The six-rowed barley selected populations were grown under two different rainfed conditions at Izraa Agricultural Experiment Station of ACSAD, Daraa governorate, Syria and Kafrdan Agricultural Experiment Station of ACSAD, Beqaa governorate, Lebanon for three successive seasons (2016/17, 2017/18 and 2018/19) to compare the efficiency of four selection methods. Results of analysis of variance and combined analysis showed significant and/or highly significant differences among lines derived from the ten populations and the four methods of selection for all studied traits and the interaction between selection methods x genotypes were highly significant for all studied traits. Mean performance of selected genotypes for the two selection methods SSDM and PM recorded the highest values in the ten populations and were significantly superior to BM and MBM methods. MBM and the SSDM recorded the highest values of phenotypic variance (σ2p) and genotypic variance (σ2g), respectively, for all studied traits as well as phenotypic variance in SSDM for no. of grains/spike, 1000 grain weight, grain yield/plant and straw yield/plant under both locations in addition to plant height and no. of spikes/plant under Kafrdan conditions. While, Broad sense heritability estimates ranged from intermediate to high according to selection method and population in all studied traits. Meanwhile, values of genetic advance and genetic gain were lower in the two methods of selection BM and MBM compared to those of PM and SSDM for all traits under Izraa and Kafrdan conditions. Results indicated that visual selection for yield by PM and/or SSDM seemed to be an effective under both locations than the other two methods and selected promising lines from the three populations 3, 5 and 7 at early generation for testing in barley could led to improved higher yielding and more tolerant lines under semiarid conditions.