Nitrogen is the nutrient that most influences productivity and pasture quality but nitrogen fertilization efficiency still needs to be increased. The objective of this work was to study the efficiency of nitrogen sources and doses (N) on the yield of Panicum maximum cv. Mombasa applied during rainy and dry periods. The soil of the experimental area is an ultisol with a sandy texture. The experimental design was a randomized block in a factorial scheme with four replications, two sources of nitrogen: urea and ammonium nitrate and five N doses (0, 50, 100, 150 and 200 kg ha -1 per cutting), in total 0, 300, 600, 900 and 1200 kg ha -1 per season-dry or rainy, respectively). The dry matter yield (DMY) and nitrogen use efficiency (NUE), agronomic efficiency (AE), recovery of applied nitrogen (RAN) and physiological efficiency (PE) were evaluated. The results showed that N source did not influence DMY, NUE, AE, RAN and PE. However, the N doses influenced DMY positively, and negatively NUE, AE, RAN and PE. Generally, the increase was occurred in DMY with the increase of N doses and decrease in NUE, AE, RAN and PE. As the N source did not influence the analyzed attributes, it is recommended to use urea, because it is a fertilizer with the highest N concentration and lowest cost per unit of nutrient, at a dose of 100 kg ha -1 per cut, favoring DMY gain. In addition this urea assists with the management and maintenance of the Mombasa guineagrass.