Nitrogen deposition in the northeastern US changed N availability in the latter part of the twentieth century, with potential legacy effects. However, long-term N cycle measurements are scarce. N isotopes in tree rings have been used as an indicator of N availability through time, but there is little verification of whether species differ in the strength of this signal. Using long-term records at the Fernow Experimental Forest in West Virginia, we examined the relationship between soil conditions, including net nitrification rates, and wood δN in 2014, and tested the strength of correlation between tree ring δN of four species and stream water NO loss from 1971 to 2000. Higher soil NO was weakly associated with higher wood δN across species, and higher soil net nitrification rates were associated with higher δN for Quercus rubra only. The δN of Liriodendron tulipifera and Q. rubra, but neither Fagus grandifolia nor Prunus serotina, was correlated with stream water NO. L. tulipifera tree ring δN had a stronger association with stream water NO than Q. rubra. Overall, we found only limited evidence of a relationship between soil N cycling and tree ring δN, with a strong correlation between the wood δN and NO leaching loss through time for one of four species. Tree species differ in their ability to preserve legacies of N cycling in tree ring δN, and given the weak relationships between contemporary wood δN and soil N cycle measurements, caution is warranted when using wood δN to infer changes in the N cycle.