Abstract: A long-term experiment was carried out on a typical loess soil in Hequ county, northeastern Shanxi province, to study the soil N fertility. Results showed that soil nitrogen content in the experiment area are much lower than those in other similar areas and the nitrogen supplying potential of soils is low (total N 0.4 g/kg averaged and alkali-hydrolyzable N 16.76 mg/kg averaged). The amounts of soil nitrogen content in arable land were highest, followed by pasture land and forest land. After 13 years continual application of single organic fertilizer or organic fertilization incorporated with chemical fertilizer, total nitrogen content increased by 12.0%105.6% in the top soil (020cm). But no significant differences of total N contents were found in the subsurface layer (2040cm) in all treatments. The alkali-hydrolyzable nitrogen contents showed a rising tendency with an increase by 32%299% for each treatment in the top soil which could be attributed to the successive application of organic fertilizer. Soil N was surplus in millet field and was roughly balance in potato fields for all fertilizer applied treatments. N application rate of current crop should be confirmed in terms of N requirement of current crop and the yield of previous crop in the rotation system. As to millet and potato, fertilization should emphasis the N fertilizer and organic fertilizer. The rational application ration of N to P for millet should be 1 to 0.50.7 and 1 to 0.7 for potato.