In Brazil, the livestock systems are mainly forage‐based, and nitrogen inputs are an uncommon practice, despite its great benefits. The objective was to evaluate the effect of nitrogen input on the physiological responses of Marandu palisadegrass (Urochloa brizantha [Hochst. ex A. Rich.] R. D. Webster) and Mombaça guineagrass (Megathyrsus maximus [Jacq.] B. K. Simon & S. W. L. Jacobs), and its impacts on the forage accumulation rate (FAR). The experiment was carried out in Sinop, MT, Brazil, in a randomized block design with a 2 × 2 × 5 factorial arrangement (two cultivars: Mombaça and Marandu; unfertilized and fertilized [50 kg N ha−1 cycle−1]; and five seasons: autumn 2015, winter 2015, spring 2015, summer 2015 and autumn 2016) with three replicates. The experimental period was from March 2015 to June 2016, when plots were harvested to mimic intermittent stocking. Fertilized Mombaça presented greater FAR pastures (74.0 kg DM ha−1 day−1) than unfertilized (26.6 kg DM ha−1 day−1) and Marandu fertilized (52.5 kg DM ha−1 day−1), with the greatest values during the summer (147.6 kg DM ha−1 day−1). The greatest leaf (A) and canopy photosynthesis (CP) rates occurred on fertilized pastures during the summer 2015 and autumn 2015. Marandu presented 19% greater A than Mombaça. During the winter, under water stress, Marandu had greater water use efficiency (WUE). However, during the summer 2015 and autumn 16, Mombaça had the greatest WUE (12.76 μmol CO2/mol H2O). In all seasons, Marandu presented the greatest chlorophyll index, which may support the greatest photosynthetic rates. It was concluded that Mombaça was highly responsive to nitrogen input during the rainy season with accumulation over 140 kg DM ha−1 day−1 and is recommended to intensified systems. Marandu had the greatest WUE in the lower precipitation seasons and is an alternative to a year‐round strategy in forage‐based systems.