Relations among nitrogen load, soil acidification and forest growth have been evaluated based on short-term (<15 years) experiments, or on surveys across gradients of N deposition that may also include variations in edaphic conditions and other pollutants, which confound the interpretation of effects of N per se. We report effects on trees and soils in a uniquely long-term (30 years) experiment with annual N loading on an un-polluted boreal forest. Ammonium nitrate was added to replicated (N=3) 0.09 ha plots at two doses, N1 and N2, 34 and 68 kg N ha−1 yr−1, respectively. A third treatment, N3, 108 kg N ha−1 yr−1, was terminated after 20 years, allowing assessment of recovery during 10 years. Tree growth initially responded positively to all N treatments, but the longer term response was highly rate dependent with no gain in N3, a gain of 50 m3 ha−1 stemwood in N2 and a gain of 100 m3 ha−1 stemwood in excess of the control (N0) in N1. High N treatments caused losses of up to 70% of exchangeable base cations (Ca2+, Mg2+, K+) in the mineral soil, along with decreases in pH and increases in exchangeable Al3+. In contrast, the organic mor-layer (forest floor) in the N-treated plots had similar amounts per hectare of exchangeable base cations as in the N0 treatment. Magnesium was even higher in the mor of N-treated plots, providing evidence of up-lift by the trees from the mineral soil. Tree growth did not correlate with the soil Ca/Al ratio (a suggested predictor of effects of soil acidity on tree growth). A boron deficiency occurred on N-treated plots, but was corrected at an early stage. Extractable NH4+ and NO3−were high in mor and mineral soils of on-going N treatments, while NH4+ was elevated in the mor only in N3 plots. Ten years after termination of N addition in the N3 treatment, the pH had increased significantly in the mineral soil; there were also tendencies of higher soil base status and concentrations of base cations in the foliage. Our data suggest the recovery of soil chemical properties, notably pH, may be quicker after removal of the N-load than predicted. Our long-term experiment demonstrated the fundamental importance of the rate of N application relative to the total amount of N applied, in particular with regard to tree growth and C sequestration. Hence, experiments adding high doses of N over short periods do not mimic the long-term effects of N deposition at lower rates.