Abstract:[Objective] The effect of plant growth-promoting rhizobacteria (PGPR) on bacterial wilt is unstable, which poses a challenge to the application of beneficial microorganisms. Thus, developing a stable and efficient PGPR flora against bacterial wilt is the key to the biocontrol. [Methods] The 8 anti-Ralstonia solanacearum (Rs) strains of PGPR (112, 114, Ba-S, TLZZ, LX4, LX7, Ps-S, and VC110) that had been screened out by our research group and Rs were studied. Based on the composition of tobacco root exudates, the microbial restriction microsystem, microplate, green fluorescent protein labeling, quantitative PCR, rod model design, and other methods were employed to explore the mechanism of PGPR against Rs in tobacco rhizosphere. In addition, experiment on the disease-resistant and growth-promoting effects was carried out in the field. [Results] LX4, Ba-S, and LX7 can make full use of amino acids and carbohydrates in tobacco root exudates as carbon sources to inhibit the growth of Rs. LX7 and 112 suppressed Rs with all acid carbon sources, and the highest antibacterial rates were 40.12% (LX7+lactic acid) and 35.15% (112+citric acid), respectively. The basal niche breadth (Bsw) of Ps-S, 112, and VC110 was 41.9%, 41.0%, and 38.1% higher than that of Rs (Bsw=2.56), respectively. The basal niche overlap index (NOI) of Ba-S was 4.88%-61.76% significantly higher than that of any other PGPR strains, and this strain obviously competed with Rs for nutrients. The average disease index was 27.01% and the average count of Rs in tobacco rhizosphere was 1.1×104 CFU/g in the treatments with the combinations of four PGPR strains, which were significantly lower than those in the treatments with one PGPR strain and two strains. The tobacco plants treated with 8 strains of PGPR were free from the wilt, with the average Rs count of 3.5×102 CFU/g. The utilization efficiency of different carbon sources by combination 32 (consisting of LX4, Ba-S, LX7 and VC110) was significantly higher than that by Rs and particularly the utilization efficiency of alcohols and carbohydrate by the combination was 1.62 and 1.41 folds that by Rs. The field effect of combination 32 against Rs was significantly stronger than that of other treatments. Especially, the Rs-controlling rate was 27.18%, 60.05%, and 54.80% higher than that of combinations 39 (Ba-S, VC110, 114, and TLZZ), 40 (LX4, LX7, VC110, and 112), and 43 (Ba-S, VC110, 114, and 112), respectively. The yield and output value in the treatment with combination 32 were the highest among all treatments, which were 67.50% and 73.53% higher than those of the control treatment, separately. [Conclusion] Via the nutrient competition or antagonistic characteristics, different PGPR strains can be fully utilized to develop PGPR flora against R. solanacearum. The diverse PGPR floras have significantly stronger ability to resist the invasion of pathogenic bacteria.