Ecosystem retention of N is mediated by interactions among plant, soil, and microbial processes. These are likely to change with forest ecosystem development as living plant biomass accumulation slows and detrital biomass increases. We investigated linkages among N storage, N cycling processes, and N leaching losses in a study of replicate mid- successional (80-yr-old) and late-successional (uneven-aged old-growth). northern hardwood forests in the western upper peninsula of Michigan, USA. Our study tested hypotheses that detrital biomass and microbial immobilization of N function as larger N sinks and corre- spond to greater N retention in old-growth compared to maturing second-growth forests. Aboveground living and detrital biomass pools were greater in old-growth compared to second-growth forest, a difference due largely to coarse woody debris (CWD). The total amount of N in detrital pools was significantly greater in old-growth than second-growth forests. We also found more rapid rates of microbial N immobilization in old-growth forests than in second-growth forests. In situ net N mineralization ranged widely among individual stands and did not differ between old- and second-growth forests. Nitrogen (organic + inorganic) leaching did not differ significantly between old-growth and second-growth forests, and was substantially lower than wet deposition inputs. Nitrate leaching losses were significantly related to soil NO3- pools, litterfall N flux, and fine root biomass across both old- and second-growth forest stands. We conclude that CWD and microbial N uptake and turnover are greater N sinks in old-growth than in second-growth forests. This apparent N sink was not the primary factor influencing N leaching loss, however. Patterns of N dynamics among individual forest ecosystems indicate that N losses correspond to net rates of N mineralization and to litterfall N flux (indicators of plant N cycling), which are independent of forest age, biomass pools, and gross N transformations at the successional stages that we compared.