Not all nitrogen (N) inputs have been accounted for in forested ecosystems. We sought to account for N2 fixation and dry deposition using a lysimeter mass—balance approach. Large sand—filled, field lysimeters were used to construct 5—yr nitrogen budgets for two N2—fixing trees, two pines, and a nonvegetated control soil. This approach is a promising and straightforward technique for quantifying otherwise difficult—to—measure fluxes. Accurate assessment of changes in N storage combined with direct measurement of N inputs in precipitation and losses from leaching allowed as to estimate fluxes. Gains of N in pine systems were greatest in vegetation and litter, overshadowing combined losses from mineral soil and leaching by about threefold. Rapid acetylene reduction in pine rhizospheres and in cultures from washed roots suggests that unexplained gains are due to associative N2 fixation. These results provide strong evidence for N2 fixation in pine systems of °50 kg°ha—1°yr—1 N. The symbiotic N2—fixing trees black locust and black alder fixed 2 and 5 times more N2, respectively, than did pines. In all systems, input in precipitation and dry deposition were relatively unimportant to the N budget. Unexplained losses of N from the nonvegetated control suggest that denitrification is an important flux. Mineral soil organic matter declined sharply and significantly in pines (20%) and even more so in the nonvegetated control (40%). Symbiotic N2—fixing trees caused a small, nonsignificant increase in mineral soil organic matter and large, significant increases in litter layer organic matter and large, significant increases in litter layer organic matter. Bulk density (0—20 cm) declined by 5% under symbiotic N2—fixing trees and increased by 5% in one pine sandbox. Correction for soil expansion or collapse did not greatly alter estimates of unexplained N or N2 fixation. Pines with rhizospheres that fix N2 at the rates we observed might be used to restore degraded land and to create silvicultural systems that are N self sufficient. We first need to better understand the microbiology, tree genetics, and soil conditions that lead to rapid N2 fixation in pine ecosystems.