【Objective】Nitrogen is the most important nutrient for plant growth, and so an important index for evaluation of soil fertility, too. Spatio-temporal variation of soil nitrogen can cause global nitrogen shifting between "source" and "sink", thus affecting nitrogen recycling in the terrestrial ecosystem. The study on spatio-temporal variation of soil total nitrogen (TN) may help effectively reveal relationships between spatio-temporal patterns of soil TN and ecological processes, which is of great significance to sustainable utilization of the soil resources. As in agricultural regions, the soil is subjected to dual influences, natural and artificial, the study on spatio-temporal variation of TN in the soil may help expose more accurately how artificial and natural factors affect the variation and guide rational fertilization and sustainable utilization of the soil in the regions. 【Method】 To that end, analysis was done of the data of the 45 profiles of the second national soil survey in 1985 and the 61 profile of the soil survey in 2015 in the Huangshui River Basin of Qinghai Province, using the classical statistics including descriptive statistics analysis, correlation analysis and variance analysis, and the geostatistics including semi-variance analysis and Ordinary Kriging analysis, for spatio-temporal variations of TN in the 0~15 cm and 15~30 cm soil layers and their influencing factors in the past 30 years. 【Result】Results show that: by excluding abnormalities, the data valid for analysis consisted of 45 items of the 0~15 cm soil layer and 44 items of the 15~30 cm soil layer obtained in 1985 and 61 items of the 0~15 cm soil layer and 58 items of the 15~30 cm soil layer obtained in 2015. From 1985 to 2015, the mean TN content in the 0~15 cm and 15~30 cm soil layers increased from 1.33 g kg^-1 to 2.25 g kg^-1 and from 1.15 g kg^-1 to 2.08 g kg^-1, or by 69.2% and 80.87%, respectively, and variation coefficient of the mean soil TN increased from 58.65% to 68.0% and from 65.22% to 67.79%, respectively, which indicates that soil TN content during the period varied at a medium level; Semi-variance analysis shows that the TN data of the two soil layers in 1985 fit the Exponential model, while those of the 0~15 cm soil layer and the 15~30 cm soil layer in 2005 fit the Gaussian model, and the Spherical model, respectively. Nugget coefficients of the TN data of the two soil layers in 1985 were both <25%, exhibiting strong spatial correlations, which indicates that spatial distribution of the soil TN was mainly affected by structural factors such as topography, parent material, temperature and precipitation, while those in 2015, both varied between 25% and 75%, displaying medium spatial autocorrelations, which indicates that spatial distribution of the soil TN during this period was governed by both structural factors and random factors, simultaneously. The change in spatial distribution autocorrelation from strong in 1985 to medium in 2015 suggests that the influences of random factors like human activities on soil TN was strengthening in the region, thus weakening the spatial distribution autocorrelation in recent years; In the past 30 years, soil TN content in the Basin was on a rising trend, demonstrated mainly in the western, northern and central parts of the region while in the southeast part it was on a reverse trend region decreased. A first-falling-and-then-rising trend was found In 1985 along the east-west and north-south directions, and in 2015 a holistically rising one was along the southeast-northwest direction. The graph of spatial distribution of standard deviations of TN contentprediction demonstrates that the predicted spatial distribution map of soil TN based on Ordinary Kriging of the two lots of TN data is good enough to reflect the real situation Natural factors, such as climate, elevation, soil texture, soil type, and so on and artificial factors, like land use, fertilization, planting structure, agricultural mechanization, etc. acted jointly causing sharp spatio-temporal variation of soil TN, and the intervention of human beings in soil nutrients exhibited an increasing trend. 【Conclusion】Considering the high spatio-temporal variability of TN content, it is suggested that a long-term soil TN monitoring system be set up in the Huangshui River Basin, and that partition management of agricultural nitrogen fertilization should be enforced, so as to ensure sustainable development of the agriculture of the whole Basin.