This experiment was carried out at Embrapa-Centro Nacional de Pesquisa de G
ado de Leite, located in Juiz de Fora, MG, Brazil, from October, 1995 to Oc
tober, 1997, to evaluate the effects of five N doses (0, 100, 200, 400 and
600 kg/ha/year) and three cutting intervals (two, four and six weeks, durin
g the rainy season, and four, six and eight weeks, during the dry season) o
n dry matter production and crude protein content of tifton 85. Data were a
nalyzed as randomized blocks design in split-plot, with three replicates. T
he annual dry matter yield increased up to the highest dose of 600 kg/ha/ye
ar of N fertilization and up to four weeks cutting interval, during rainy s
eason, and six-week cutting intervals, during the dry season. The highest p
roduction of dry matter was 23.1 ton/ha/year, of which 17.8 ton/ha were obt
ained in the rainy season. The lowest production was 2.6 ton/ha/year, of wh
ich 1.9 ton/ha were obtained in the rainy season. Crude protein content var
ied inversely with cutting intervals and as the amount of N fertilization i
ncreased up to dose of 600 kg/ha/year. The crude protein contents did not d
iffer significantly with cutting intervals during the raining season. The h
ighest level of crude protein obtained was 21.7%. Tifton 85 persistency was
hampered in the absence of N fertilization and the shortest cutting interv
al.