In order to elucidate the intensity and mechanism of the impact of atmospherically derived nitrate on the soil nitrogen dynamics, dual isotope measurements on nitrate (δ15N and δ18O) in soil waters were conducted in a forested watershed in central Japan. Soil waters have been sampled from various horizons of the soil profiles to elucidate the vertical distributions of nitrogen transformation processes. In general, the δ18O of nitrate in rainwater is remarkably higher than that produced by nitrifying bacteria in soils. Accordingly, the δ18O can often be used as an index of the impact of the atmospherically derived nitrate. While soil waters in <20cm depth had a strong signal of the atmospheric nitrate, the δ18O-nitrate in soil water decreased in the deeper soil horizons, indicating that the dominant source of nitrate in this soil profile was nitrification. The net nitrate production of this soil profile was about 18 kg-N ha-1year-1, and deposited nitrate was about 6 kg-N ha-1year-1. Assuming that the annual mean δ18O of deposited nitrate was 60 ⁰/₀₀, and the mean value of bacterially produced nitrate in soil was about 0 ⁰/₀₀, the average value for soil nitrate pool could be ~15 ⁰/₀₀. However, the observed δ18O of the soil and groundwater was 0 to 6 ⁰/₀₀ and remarkably smaller than the above estimation based on annual mass balance. This suggests that the gross nitrification was sufficiently higher than net nitrification rate, and the major portion of nitrate produced in soil was reused by microbes. Previous studies in temperate forests have shown from laboratory experiments using isotope dilution method that gross nitrification in soils was several times greater than net nitrification. We could verify this phenomenon from our δ18O-nitrate measurements of the field samples.