Two quantitative trait loci (QTL) from Lycopersicon hirsutum, Rcm 2.0 and Rcm 5.1, control resistance to Clavibacter michiganensis subsp. michiganensis ( Cmm). To precisely map both loci, we applied interval mapping techniques to 1,056 individuals in three populations exhibiting F(2) segregation. Based on a 1-LOD confidence interval, Rcm 2.0 mapped to a 14.9-cM interval on chromosome 2 and accounted for 25.7-34.0% of the phenotypic variation in disease severity. Rcm 5.1 mapped to a 4.3-cM interval on chromosome 5 and accounted for 25.8-27.9% of the phenotypic variation. Progeny testing of recombinant plants narrowed the QTL location for Rcm 2.0 to a 4.4-cM interval between TG537-TG091 and to a 2.2-cM interval between CT202-TG358 for Rcm 5.1. A population of 750 individuals exhibiting F(2) segregation was used to detect epistasis between both loci using ANOVA and orthogonal contrasts ( P=0.027), suggesting that resistance was determined by additive gene action and an additive-by-additive epistatic interaction. A partial diallel mating design was used to confirm epistasis, advance superior genotypes, randomize genetic backgrounds, and create recombination opportunities. This crossing scheme created a more balanced population ( n=112) containing the nine F(2) genotypic classes. Parents in the diallel were selected from the previous population based on resistance, genotype at the Rcm 2.0 and Rcm 5.1 loci, and horticultural traits. A replicated trial using the diallel population confirmed additive-by-additive epistasis ( P<0.0001). These results validate the gene action, intra -locus interaction, and map position of two loci controlling resistance to Cmm.