Genome-wide association studies have identified many variants associated with complex traits, but in general these variants explain only a small fraction of the estimated heritability of the trait. Several hypotheses have been proposed to explain this “missing heritability.” Complex traits could be caused by rare undiscovered variants with relatively large effect sizes, by the additive action of many individually small-effect variants, or by significant gene-by-gene or gene-by-environment interactions that modify otherwise small-effect variants. We are using the yeast mating pathway as a model complex trait to experimentally assess the relative impact of additive and epistatic genetic variation. Yeast mating is controlled by a well-characterized pheromone-responsive MAP kinase pathway. We used a technique called deep mutational scanning to profile the effect on mating of thousands of variants in three mating pathway genes: STE5 (scaffold protein), STE7 (MAP kinase kinase) and STE12 (transcription factor). We transformed yeast null for one of these genes with a library of ~100,000 plasmids, each carrying a variant of the gene. We mated the transformants, selected for diploids, and sequenced the population of variants before and after mating by high-throughput sequencing. Mutations that impair gene function are depleted in the population after mating, while mutations that allow a wild-type level of mating are enriched. From these experiments, we identified variants in all three genes with the full spectrum of effect sizes, from null to wild-type. In addition, we tested the mating ability of the same population of variants in the presence of an Hsp90 inhibitor and at high temperature. Hsp90 is a chaperone protein that buffers the effects of mutations in many genes. From these data, we will identify variants that are sensitive to a known genetic modifier and the environment (high temperature), and determine the prevalence of buffered genetic variation in these genes. Finally, we identified ~600 small-effect variants in each gene, and in future experiments we will combine these small-effect variants in two mating pathway genes at a time. This experiment will allow us to assess simultaneously the degree of epistasis between tens of thousands of variants in both physically interacting and non-interacting proteins.