A simulation model for the prediction of grassland (Lolium perenne) productivity under conditions of climate change is described and validated for grass growing in the Wageningen Rhizolab, Wageningen, The Netherlands. In this work the model was used to study the impact of different management strategies on the productivity of grassland under present and increased atmospheric CO2 concentrations. In LINGRA-CC simulated key processes are light utilization, leaf formation, leaf elongation, tillering and carbon partitioning. The daily growth rate is determined by the minimum of a sink and a source term. As in a previous model (LINGRA), the potential growth of the sink depends on the mean daily temperature, and can be modified by the effects of the availability of assimilates on tillering. The growth of roots is calculated from the amount of carbohydrates the shoot is unable to utilize when the number or activity of the sinks is small (overflow hypothesis). The main difference between LINGRA and LINGRA-CC is the way the source of assimilates for growth is calculated. Assimilate production depends on intercepted radiation, and a photosynthetic light-use efficiency (LUE) calculated as a function of CO2, temperature, light intensity and the Rubisco concentration of upper leaves. Other differences are that in LINGRA-CC, the specific shoot area for new growth depends on the level of reserves. Data from two independent experiments with L. perenne swards, grown in enclosures at two levels of CO2 during 1994 and 1995, were used to calibrate and validate the model, respectively. The model predicted well the observed amounts of harvested biomass, and the dynamics of the leaf area index, tiller number and specific shoot area. LINGRA-CC was used to study the effects of different combinations of cutting interval and cutting height on biomass production, at ambient (350 μmol mol−1 CO2) and double (700 μmol mol−1 CO2 ) CO2 conditions. Under both ambient and doubled CO2, maximum biomass was produced with cuttings of leaf area index >1, and at cutting intervals of 20 and 17 d for ambient and increased CO2 environments, respectively. Under high CO2 conditions the cutting interval for maximum yield was 15% shorter than at ambient CO2. However, the gain in harvested biomass obtained by reducing the cutting interval by 3 d under high CO2 conditions was negligible.