Grain yield and plant height of 80 recombinant chromosome substitution lines (RCSLs) of barley were measured in six environments with contrasting available moisture profiles. Two environments were in OR, USA (Moro and Pendleton) during one growing season (2004), and four in Chile (Cauquenes and Santa Rosa) during two growing seasons (2004/2005 and 2007/2008). From the yield data obtained in the different environments, yield adaptability (Finlay-Wilkinson slope) and stability (deviations from regression) were calculated. Two commercial cultivars (Harrington and Baronesse) were used as checks in all environments. Marker-quantitative trait associations were identified using 47 simple sequence repeats (SSRs) and the general linear model (GLM) implemented in TASSEL. The mean plant height and grain yield of the 80 RCSLs differed greatly across environments, reflecting differences in water availability. In all environments, there were significant differences (P < 0.05) in grain yield among RCSLs. There was also abundant variation in yield adaptability, indicating a differential response of the RCSLs to environmental conditions across environments. Using principal component analysis, it was possible to identify genotypes with better agronomic performance than the recurrent parent cv. Harrington. The association analysis revealed 21 chromosomal regions that were highly correlated with differences in grain yield, plant height and/or yield adaptability (Finlay-Wilkinson slope). In approximately one-fourth of the cases, the H. spontaneum donor contributed favorable alleles. The associations were referenced to the quantitative trait loci (QTL) for the same traits reported in the literature.