Forage and turf grasses began to diverge from each other about 50 years ago with the advent of turfgrass breeding programs. With the exception of Lolium multiflorum, forage grasses are largely undomesticated – cultivars have little or nothing to phenotypically distinguish them from wild counterparts. Most forage grass breeding programs rely heavily on germplasm collected from wild or naturalized stands, with strong reliance on national or local germplasm collections. For most species, cultivars may be developed and released after only two or three cycles of selection and recombination. Core collections, accession evaluation data, and accession passport data may be particularly valuable in some of these forage grasses. In contrast, numerous cycles of selection and breeding have been completed for most turfgrasses, resulting in distinct phenotypic differentiation from wild counterparts of these species. This differentiation was observed as early as the beginning of the 20th century, with the identification of high-quality and pest resistant clones in old turf sods. Collection, recombination, and continued selection within populations derived from some of these clones have resulted in elite germplasm that can be easily distinguished from wild germplasm of most turfgrass species. As a result of these successes, turfgrass breeders continue to rely heavily on continued collection of desirable clones from old turf sods. All private and most public turfgrass breeding programs develop and maintain their own germplasm collections that are not represented in national or public gene banks. Clones collected overseas are converted to seed prior to introduction into the USA, using overseas collaborators and phytosanitary inspections. Following introduction, intensive selection and recombination are generally emphasized over preservation and maintenance of the original accessions. The highly heterozygous and heterogeneous nature of turfgrasses helps to preserve a large number of genes in cultivated germplasm pools, but intensive selection probably acts to eliminate many rare alleles.