Dissecting the genetic architecture of quantitative traits is a formidable challenge in modern biology. Genetic resource populations are a powerful tool for understanding the contributions of alleles segregating in natural populations to variation in quantitative traits. We present the “Yeast Diversity Mapping Panel”, a genetically diverse panel of nearly 800 F2 segregants from crosses between wild, laboratory, and domesticated strains of Saccharomyces cerevisiae. This genetic resource population has been carefully designed to capture the bulk of genetic diversity present in budding yeast, maximize allelic richness across the panel, and incorporate ancestry from parental strains that show substantial variability across >250 phenotypic axes. We describe a novel method for computationally deconvolving haplotype blocks from pooled, unbarcoded libraries of genetically diverse segregants, and demonstrate the use of this technique to obtain full genome sequence information for our mapping panel. The resulting high-resolution genotypes can be integrated with patterns of allelic effects to narrow QTL intervals, and will facilitate identification of candidate causal alleles for functional follow-up. We demonstrate the utility of this mapping panel by characterizing the genetic architecture of a quantitatively varying phenotype, growth response to osmotic stress. Overall, this panel constitutes a powerful resource for interrogating the genetic basis of complex trait variation, and will facilitate studies of dynamically variable phenotypes, epistasis, and gene-environment interaction.