The responses of temperate and boreal forest ecosystems to increased nitrogen (N) inputs have been varied, and the responses of soil N pools have been difficult to measure. In this study, fractions and pool sizes of N were determined in the forest floor of red spruce stands at four sites in the northeastern U.S. to evaluate the effect of increased N inputs on forest floor N. Two of the stands received 100 kg N ha-1 yr-1 for three years, one stand received 34 kg N ha-1 yr-1 for six years, and the remaining stand received only ambient N inputs. No differences in total N content or N fractions were measured in samples of the Oie and Oa horizons between treated and control plots in the three sites that received N amendments. The predominant N fraction in these samples was amino acid N (31-45 % of total N), followed by hydrolyzable unidentified N (16-31% of total N), acid- soluble N (18-22 % of total N), and NH4/+-N (9-13 % of total N). Rates of atmospheric deposition varied greatly among the four stands. Ammonium N and amino acid N concentrations in the Oie horizon were positively related to wet N deposition, with respective r2 values of 0.92 and 0.94 (n = 4, p < 0.05). These relationships were somewhat stronger than that observed between atmospheric wet N deposition and total N content of the forest floor, suggesting that these pools retain atmospherically deposited N. The NH4/+- N pool may represent atmospherically deposited N that is incorporated into organic matter, whereas the amino acid N pool could result from microbial immobilization of atmospheric N inputs. The response of forest floor N pools to applications of N may be masked, possibly by the large soil N pool, which has been increased by the long-term input of N from atmospheric deposition, thereby overwhelming the short-term treatments.