Genetic mapping methods provide a unique opportunity to study the interactions of differentiated genes and genomes in a hybrid genetic background. After a brief discussion of theoretical and analytical concerns, we review the application of these methods to a wide range of evolutionary issues. Map-based studies of experimental hybrids indicate that most postzygotic reproductive barriers in plants are polygenic and that the expression of extreme or novel traits in segregating hybrids (transgressive segregation) results from the complementary action of divergent parental alleles. However, genetic studies of hybrid vigor do not concur in their interpretations of the relative roles of dominance, overdominance, and epistasis. Map-based studies of natural hybrids are much rarer, but the few existing studies confirm the polygenic basis of postzygotic barriers and demonstrate the utility of genetic linkage for detecting cryptic introgression. In addition, studies of experimental and natural hybrid lineages provide compelling evidence that homoploid hybrid speciation has occurred in nature, and that it represents a rapid and repeatable mode of speciation. Data further indicate that this mode is facilitated by strong fertility selection and high chromosomal mutation rates. We recommend that future studies of hybrid genomes focus on natural hybrids, not only because of the paucity of data in this area, but also because of the availability of highly recombinant hybrid genotypes in hybrid zones. Of particular value will be studies of long-lived or difficult-to-propagate organisms, which previously have not been amenable to genetic study.