The effect of the method of conservation of forage on endogenous N (EN) secretion was studied using a (15)N isotope dilution technique in 4 lactating Holstein cows selected from a replicated 3 x 3 Latin square. Cows were equipped with ruminal, duodenal (n = 4), and ileal (n = 2) cannulas. Diets comprised 44% concentrate plus first-cut timothy conserved either as hay or as restrictively (formic) or extensively (inoc) fermented silage. Crude protein contents of hay, formic, and inoc averaged 10.4, 13.6, and 14.8%, respectively. Total EN flow and free EN at the duodenum were increased with hay compared with silages but were similar when expressed as proportion of duodenal N flow, with total EN flow averaging 25.8, 23.9, and 23.9% for hay, formic, and inoc, respectively, and free EN at the duodenum averaging 11.5, 9.8, and 9.7% for hay, formic, and inoc, respectively. Flow of bacterial N at the duodenum originating from an endogenous source tended to be higher with inoc compared with formic. Overall, the proportion of bacterial N derived from endogenous sources and urea was similar between treatments, averaging 23 and 15%, respectively. In the feces, flow of EN was similar across treatments and averaged 31% of total fecal N. More than 70% of fecal EN originated from undigested secretions into the forestomach. Absorption of N from the forestomach tended to increase for silages compared with hay. In conclusion, EN represented an important fraction of N flowing at the duodenum and in the feces. The free EN and the total EN at the duodenum were altered by the different methods of forage conservation studied. Estimation of true dietary N supply and requirements of the dairy cow should allow for endogenous N flows and losses.