The existence of large population of ineffective native rhizobia and inconsistent performance of exotic strains in Ghanaian soils necessitate the need to identify effective and locally adapted elite strains for enhanced legume-rhizobium symbiosis. This study was designed to test the suitability of two previously selected potential elite Bradyrhizobium strains for use as inoculants on cowpea (Vigna unguiculata L. Walp) and groundnut (Arachis hypogaea L.) in multi-location experiments. Field experiments were set up at 26 locations (12 planted with cowpea and 14 planted with groundnut) in the Northern region of Ghana. Four treatments were applied at each location: inoculation with Bradyrhizobium strains KNUST 1002 and KNUST 1006, a positive nitrogen (+N) control and a negative control (without nitrogen or inoculation) arranged in randomized complete blocks with four replications. The results showed that inoculation with strains KNUST 1002 and KNUST 1006 promoted significant increases in grain yields of both cowpea and groundnut. On average, inoculating cowpea with strains KNUST 1002 and KNUST 1006 resulted in 63 and 52% increases in grain yield when compared to the negative control. Pod yields of groundnut on the other hand, were significantly (p< 0.05) increased at 57% of the study location with one or both test strains. Responses to inoculation were, however, highly variable across the different study locations (i.e. significant Treatment ⸼ Location Interaction, TLI). A detailed analysis of this significant TLI based on the genotype main effect (G) plus genotype-by-environment (GE) interaction (GGE) biplot analysis revealed that location contributed 71 and 88% of the variation observed in cowpea and groundnut, respectively, and grouped the locations into mega-environments. These results demonstrate that native elite Bradyrhizobium strains KNUST 1002 and KNUST 1006 have potential for use as inoculants to increase cowpea and groundnut production in Northern Ghana.