Forage-based livestock systems are complex and interactions among animals, plants, and the environment exist at several levels of complexity, which can be evaluated using computer modeling. Pastures are key to livestock production in Brazil because they make for low feeding costs and, more recently, have been regarded to promote higher marketability to the final animal products. Despite the importance of grasslands for livestock production in Brazil, tools that assist producers to make decisions in forage-livestock systems are scarce. The objective of this research was to use the CROPGRO – Perennial Forage Model to simulate accurately the irrigated and rainfed growth of Marandu palisadegrass (Brachiaria brizantha cv. Marandu), the most widely grown forage in the country, using the model previously calibrated for the tall-growing Xaraes cultivar of the same species, under non-limiting water conditions. Our null hypothesis was that the forage model previously calibrated for a given cultivar of species, can accurately simulate the growth and forage yield of a new cultivar of the same species under irrigated and rainfed conditions. Data used to calibrate the model included forage production, leaf photosynthesis, leaf area index, light interception and plant nitrogen concentration from a field experiment conducted in 2011, 2012 and 2013 in Piracicaba, SP, Brazil.  Agronomic and morphological differences between the two grasses, such as maximum leaf photosynthesis, nitrogen concentration and temperature effect on growth rate, were considered in the calibration. Under rainfed conditions, the simulations using Penman-FAO method gave more realistic water stress response than using the Priestley and Taylor method. Calibration results suggest that the CROPGRO – Perennial Forage Model can be used to adequately simulate growth of Marandu palisadegrass under irrigated and rainfed conditions.