The Oryza sativa (rice) resistance gene Pia confers resistance to the blast fungus Magnaporthe oryzae carrying the AVR-Pia avirulence gene. To clone Pia, we employed a multifaceted genomics approach. First, we selected 12 R-gene analog (RGA) genes encoding nucleotide binding site-leucine rich repeats (NBS-LRRs) proteins from a region on chromosome 11 that shows linkage to Pia. By using seven rice accessions, we examined the association between Pia phenotypes and DNA polymorphisms in the 10 genes, which revealed three genes (Os11gRGA3-Os11gRGA5) exhibiting a perfect association with the Pia phenotypes. We also screened ethyl methane sulfonate (EMS)-treated mutant lines of the rice cultivar 'Sasanishiki' harboring Pia, and isolated two mutants that lost the Pia phenotype. DNA sequencing of Os11gRGA3-Os11gRGA5 from the two mutant lines identified independent mutations of major effects in Os11gRGA4. The wild-type 'Sasanishiki' allele of Os11gRGA4 (SasRGA4) complemented Pia function in both mutants, suggesting that SasRGA4 is necessary for Pia function. However, when the rice cultivar 'Himenomochi' lacking Pia was transfected with SasRGA4, the Pia phenotype was not recovered. An additional complementation study revealed that the two NBS-LRR-type R genes, SasRGA4 and SasRGA5, that are located next to each other and oriented in the opposite direction are necessary for Pia function. A population genetics analysis of SasRGA4 and SasRGA5 suggests that the two genes are under long-term balancing selection.