The major setback in successfully obtaining an effective inoculant is overcoming difficulties in formulating a viable and user-friendly final product as the live nature of the active ingredient underscores the importance of formulation in maintaining the microbial cells in a competent state. Co-cultures of rhizobia and PGPR (Plant Growth Promoting Rhizobacteria) are also known to influence the efficacy of the symbiotic bacteria on plant biological nitrogen fixation. A greenhouse experiment was set to assess the formulation effect of one strain i.e. Bradyrhizobium japonicum, 532c (granules, liquid and broth) and to determine the efficiency of co-inoculation of Bacillus subtilis with two strains of Bradyrhizobium japonicum (532c and RCR 3407). The objectives were evaluated on 2 soybean (Glycine max L.) varieties: Nyala, a non-promiscuous variety and TGx1740-2F, a promiscuous test variety. A non sterile soil from Central Kenya (Chuka) classified as a Nitisol was used. Nodule occupancy was determined by PCR-RFLP. Most of the inoculants showed increased nodulation and biomass yields as compared to the un-inoculated controls with a higher response seen in the promiscuous TGx1740-2F variety as compared to the non-promiscuous variety. The liquid and granule-based inoculants had higher biomass yields suggesting an impact of formulation on the effectiveness of the inoculants. The co-inoculants also gave higher yields but showed no significant differences to the rhizobial inoculants alone. Nodule occupancy was 100 % for all the rhizobial inoculants as well as the co-inoculants emphasizing the infectivity and high competitiveness of 532c and RCR 3407 strains even in the presence of indigenous strains (80-113 cell/g of soil). These inoculants, though not initially made for SSA countries, showed promising increased yields in a Kenyan soil containing significant populations of native rhizobia nodulating soybean, signifying a possibility of their adoption in increasing soil fertili