The majority of the economically important tropical forage grasses (Genera Panicum, Paspalum, Pennisetum, Cenchrus and others) reproduce through facultative apomixis. The apomixis fenomenus has allowed, throughout the evolution of the species, the preservation of highly heterozygous genotypes (specially interspecific hybrids) which, being highly steriles, would have been otherwise eliminated by the natural selection, had the assexual reproduction not occurred. Because of that, there is a high quantity of genetic variability "hidden" in forage grasses, which being released, could be used for the selection of new cultivars presenting very desirable chaiacteristics (percentage of protein/dry matter, flowering cycle, seed retention, cold tolerance and others) has presented litlle or no variation within populations, which turn te selection process into a difficulttask. In colonial grass (Panicum maximum Jacq.), for example, all known cultivars and ecotypes present indetenninated flowering (3-4 months), high seed shedding, relatively low protein (5-7%) and low cold tolerance. The obtention of F1 hybrids, through crossings between setected sexual lineages (females) and cultivars and/or apomilic ecotypes (mates) is the most viable explotation process of the "hidden" genetic variability, marking possibte the selection of new cultivars with one or more of the desidble characteristics mentioned, within a short period of time. Besides the obtention of superior F1 apomitic hybrids, the selection of new ecotypes and the introduction and evaluation of exotic materials also allow the explotation of variabilities already existing in the specie and, from that, high performing cultivars could also result.