The agronomic value of grasslands tends to decrease over time, leading to the need for repeated ploughing and resowing, which cause long term environmental damage. In order to extend the time during which grasslands are productive, we must understand the causes for this decline. Changes in population genetic structure of agronomic grasses due to the interaction of selection, migration and drift may provide part of the answer. Selection pressures can be mediated by abiotic factors (e.g., cultural practices, soil, climate) or biotic factors (e.g., competitors, diseases). Gene flow into sown grasslands may occur via pollen or seed from the surrounding landscape, and small effective population size may exacerbate genetic drift. Given that grasslands are composed of multiple species and cultivars, similar mechanisms may control their species composition. Different pedo-climatic conditions, different species and genetic grassland compositions and different cultural practices lead to different adaptations of grassland community. Productivity in sown grasslands will be determined initially by the sown composition of species and genotypes within species. Productivity and resistance to invasion can be enhanced by complementary resource use among species or genotypes. Declines in productivity may arise if the traits that determine the initial success of species or genotypes are negatively correlated with traits that determine productivity over the long term. If so, taking measures to enhance the maintenance of diversity over time may extend the period of high productivity in sown grasslands, and developing cultivars or mixtures for which the productivity- success trade off is unlikely, are important future goals.