Efficient Mixed-Model Association (EMMA) is a new method for correcting population structure and genetic relatedness in association mapping of model organisms. EMMA improves computational speed and reliability by leveraging the specific nature of the optimization problem in mixed models. It reduces computational complexity from cubic to linear by using spectral decomposition, enabling faster and more accurate results. EMMA was tested on inbred mouse, Arabidopsis, and maize data sets, identifying significant associations with reduced false positives. It also performed well in simulation studies, showing high statistical power when SNPs explain a large portion of phenotypic variance. EMMA's efficiency is further enhanced by using a simple genetic similarity matrix as a kinship matrix, which is positive semidefinite and effective in reducing false positives. Phylogenetic control was also introduced, using a phylogenetic tree to model genealogical history and correct for complex genetic relations. EMMA's results were consistent with previous studies, showing significant associations with known QTL and genes. An R package and webserver are publicly available for EMMA. The method is efficient, accurate, and suitable for large-scale association mapping in model organisms.Efficient Mixed-Model Association (EMMA) is a new method for correcting population structure and genetic relatedness in association mapping of model organisms. EMMA improves computational speed and reliability by leveraging the specific nature of the optimization problem in mixed models. It reduces computational complexity from cubic to linear by using spectral decomposition, enabling faster and more accurate results. EMMA was tested on inbred mouse, Arabidopsis, and maize data sets, identifying significant associations with reduced false positives. It also performed well in simulation studies, showing high statistical power when SNPs explain a large portion of phenotypic variance. EMMA's efficiency is further enhanced by using a simple genetic similarity matrix as a kinship matrix, which is positive semidefinite and effective in reducing false positives. Phylogenetic control was also introduced, using a phylogenetic tree to model genealogical history and correct for complex genetic relations. EMMA's results were consistent with previous studies, showing significant associations with known QTL and genes. An R package and webserver are publicly available for EMMA. The method is efficient, accurate, and suitable for large-scale association mapping in model organisms.