The presence of genotype-environment interactions (GE) leads to the imperfect genetic correlation between the measurements of the same trait in different environments, thereby limiting the ability of plant breeders to identify superior breeding lines or best cultivars across the environments. We analyzed genetic correlations and distributions of quantitative trait loci (QTL) controlling yields over pairs of 16 environments for a doubled haploid (DH) population of barley (Hordeum vulgare L.) derived from the cross between two six-row cultivars, Steptoe and Morex. We examined if the imperfect genetic correlation was associated with the frequency of QTLs that were concurrently detected in pairs of environments. Each of 120 environment pairs would fall into one of the four Scenarios arranged in the 2 × 2 two-way contingency table. These four Scenarios are: high (≥0.6) genetic correlation with QTL concurrence (scenario A), high genetic correlation without QTL concurrence (scenario B), low (<0.6) genetic correlation with QTL concurrence (scenario C) and low genetic correlation without QTL concurrence (scenario D). The numbers of environment pairs under scenarios A, B, C and D were 9, 8, 27 and 76, respectively. Further partitioning of the covariance due to individual concurrent QTLs confirmed the expected occurrence of scenarios A and D and it also enabled us to explain scenarios B and C. Scenario B was more likely due to the cumulative effect of undetectable concurrent QTLs with small effects and/or linked QTLs.  Scenario C was likely due to the canceling effect of concurrent QTLs and/or linked QTLs with opposite signs. This study stressed the need to jointly examine contributions of all QTLs to the magnitude of genetic correlation between environments for understanding the nature of GE.