The narrow genetic base in modern crop varieties makes them susceptible to disease epidemics and to environmental fluctuations. Over the long term, a low level of genetic variation also has a more subtle effect; it reduces the possibilities for sustained genetic improvement by plant breeders. Yet abundant genetic variation can be readily found in sexually compatible wild and exotic germplasm. Many wild ancestors of modern crop plants can still be found in their natural habitats, and seeds or propagules from those wild relatives can be collected and maintained in national and international germplasm collections. Early landraces that are not in themselves particularly prolific in modern agricultural environments provide an opportunity for breeders to expand the cultivated gene pool. We have used molecular maps and markers to identify quantitative trait loci (QTLs) in interspecific populations derived from crosses between elite, high-yielding Oryza sativa rice cultivars and the low-yielding wild ancestor, O. rufipogon. These studies have shown that a phenotypically inferior wild ancestor harbours genes that can contribute positively to rice improvement. Once identified, the genes underlying the QTLs can be cloned. This holds important implications for using naturally occurring genetic variation to amplify the genetic base of many of our cultivated varieties and contribute positively to sustainable crop production, particularly in drought and other stress-prone environments. The isolation and characterization of the genes underlying these QTLs will provide essential information about the structure, function and evolution of the target genes. As we consider new approaches to genetic manipulation of crop plants, knowledge about the range of naturally existing sequence variation at key loci and the impact of that variation on phenotypic performance will help guide our efforts. Using genomics tools and sequence information to identify useful genes, our method simultaneously paves the way for a broadening of the genetic base of cultivated crop species and provides new rationale for the conservation, characterization and use of wild and unimproved germplasm in crop improvement.