Two tall fescue [Lolium arundinaceum (Schreb.) Darbysh] forages, one an experimental host plant/endophyte association containing a novel endophyte that produces low or nil concentrations of ergot alkaloids (HM4) and the other a typical association of Kentucky 31 tall fescue and the wild-type endophyte (Neotyphodium coenophialum; E+), were autumn-stockpiled following late-summer clipping and fertilization with 56 kg/ha of N to assess N partitioning and ruminal disappearance kinetics of N for these autumn-stockpiled tall fescue forages. Beginning on December 4, 2003, sixteen 361 +/- 56.4-kg replacement dairy heifers were stratified by weight and breeding, and assigned to one of four 1.6-ha pastures (2 each of E+ and HM4) that were strip-grazed throughout the winter. Pastures were sampled before grazing was initiated (December 4), each time heifers were allowed access to a fresh pasture strip (December 26, January 15, and February 4), and when the study was terminated (February 26). Generally, fescue type and the fescue type x sampling date interaction exhibited only minor effects on total forage N, or partitioning of N within the cell solubles or the cell wall. For pregrazed forages, concentrations of N and N partitioned within the cell solubles both declined in a strongly linear relationship with sampling dates. In contrast, concentrations of cell-wall-associated N changed in erratic and often higher-ordered relationships with time, but the magnitude of these responses generally was limited. Unlike the partitioning of N within cell-wall and cell-soluble fractions, kinetic characteristics of ruminal N disappearance frequently exhibited interactions of fescue type and sampling date. For pregrazed forages, these included interactions for all response variables, and for postgrazed forages, fractions B and C, as well as rumen degradable protein. Ruminal disappearance rate for pregrazed E+ and HM4 exhibited quadratic (range = 0.057 to 0.082/h) and cubic (range = 0.057 to 0.075/h) relationships with time, respectively. For postgrazed E+ and HM4 forages, ruminal disappearance rate was unaffected (mean = 0.066/h) or only tended to be affected by sampling date (mean = 0.065/h), respectively. Concentrations of rumen degradable protein exhibited various curvilinear relationships with sampling dates, but disappearance was consistently extensive, and the overall range was relatively narrow (71.3 to 78.9% of N). These findings suggest that ruminal disappearance of N for autumn-stockpiled tall fescue forages remains extensive throughout the winter months and is only affected minimally by fescue type, sampling date, and grazing status.