Grassland production based on cutting-only and an increasing frequency of prolonged periods of drought due to climate change could lead to decreased productivity in Lolium perenne. In field experiments, we tested whether Festuca arundinacea and Phleum pratense could be suitable alternatives to L. perenne on intensively managed grassland on clay, peat and sandy soil. The three grasses were sown in mixture with Poa pratensis and Trifolium repens and subjected to different frequencies of defoliation representing a cutting-only system, simulated grazing system and a mixed system. We found that in systems with at least six defoliations, F. arundinacea represented an acceptable compromise between feed quality (6.4 MJ net energy/kg dry matter (DM), 19.3% crude protein, CP), persistence (mass proportion >90%), and DM yield (12.7 Mg/ha). However, for dairy production based on intensive cutting-only systems, the quality of F. arundinacea was insufficient (5.9 MJ net energy/kg DM, 15.6% CP). Mixtures with P. pratense as the main sown species did not differ significantly in production of net energy and CP from L. perenne in cutting-only systems on sandy soil. On peat land, all sown mixtures were invaded by Holcus lanatus. We found that under frequent defoliation conditions, H. lanatus-rich swards had comparatively good DM yields (9.2 Mg/ha) and a feed quality that would be sufficient for dairy cow nutrition (net energy, 6.2–6.4 MJ/kg DM; 18.8–20.4% CP). We conclude that there is potential to adapt the choice of grasses and mixtures in different production systems to meet the challenges of climate change.