Technical Abstract: Maize is highly genetically and phenotypically diverse. Tropical maize and teosinte are important genetic resources that harbor unique alleles not found in temperate maize hybrids. To access these resources, breeders must be able to extract favorable unique alleles from tropical maize and teosinte from their population genomic context, where they are linked with many undesired alleles that confer adaptation to tropical environments, ancient farming methods, or wild growth habit (in the case of teosinte). Long-term traditional breeding efforts have demonstrated the value of diverse germplasm to improve maize productivity, while also enhancing the genetic base of cultivated varieties. Genomics provides new opportunities to identify the genes affecting important agronomic traits and to estimate the wide range of allelic effects at such genes. New approaches to complex trait analysis, including joint multiple population analysis, genome-wide association analysis, and genomic selection, can leverage high throughput sequencing and genotyping technologies to improve our understanding of the genome-wide distribution of allele effects across the wide genetic variation in the primary gene pool of maize. Implementing this information for practical maize improvement remains a challenge.