Background: Rhizoctonia solani AG1 IA is an important pathogen of rice (Oryza sativa L.) that causes rice sheath blight (RSB). Since control of RSB by conventional measures has failed, novel strategies like application of plant growth-promoting rhizobacteria (PGPR) can be an e cient alternative.Method and Results: mRNA sequences of rice were retrieving from NCBI for candidate reference genes selction, and seven candidate reference genes (RGs), namely 18SrRNA, ACT1, GAPDH2, UBC5, RPS27, eIF4aand CYP28, were selected for their stability in real-time quantitative PCR (RT-qPCR). Different algorithms were exploited, Delta Ct, geNorm, NormFinder, BestKeeper, and Comprehensive ranking by RefFinder, to evaluate RT-qPCR of rice in tissues infected with R. solani and treated with the PGPR strains, Pseudomonas saponiphilia and Pseudomonas protegens, and potassium silicate (KSi) alone or in combination with each PGPR strain. RGs stability was affected by each treatment and treatment-speci c selection was approved and validated for nonexpressor of PR-1(NPR1) for each treatment. Conclusion: Overall, ACT1 was the most stable RG with R. solani infection alone, GAPDH2 with R. solani infection plus KSi, UBC5 with R. solani infection plus P. saponiphilia, and eIF4a with R. solani infection plus P. protegens. Both ACT1 and RPS27 were the most stable with the combination of KSi and P. saponiphilia, while PRS27 was the most stable with the combination of KSi and P. protegens 51. Pfa MW (2001) A new mathematical model for relative quanti cation in real-time RT-PCR. Nucleic acids research 29(9): e45-e45