The goal of this work was to determine the best rates of N, P2O5, and K2O to get maximum forage yield of Paspalum regnellii Bra 019186 grown in two soil fertility levels: low and medium in P and K. The experiments were set up in a typical Red-yellow latosol (Hapludox), in São Carlos, SP, Brazil. Experimental design was an incomplete (1/2)4³ factorial with a total of 32 plots, in two randomized blocks, without replication. Treatments were four rates of N and K2O (0, 100, 200, and 300 kg ha-1 in the first and second year), using urea and potassium chloride, and four rates of P2O5 ( 0, 80, 160, 240 kg ha-1 in the first year and 70, 140, 210 kg ha-1 in the second year), using triple superphosphate. Cuttings were done, mainly each 35 days, in the rain season. The N, P, K rates for maximum yield in the first year were, in kg ha-1 with low fertility: N = 210, P2O5 = 240 , K2O = 190; and with medium fertility: N = 250 , P2O5 = 240, K2O = 180, and in the second year: N = 300, P2O5 = 210, K2O = 300, in soil with low fertility, and N = 300 , P2O5 = 190, K2O = 290 in soil with medium fertility, resulting in forage yields of, respectively, 5,428 and 7,789 kg ha-1 of dry matter in the first year, and of 13,084 e 10,468 kg ha-1 in the second year.
forage yield; nitrogen; phosphorus; potassium
