In this paper, we take Beijing as a case study and employ the residential leasing parcel data from 2004 to 2015 within the Sixth Ring Road of Beijing metropolitan area. Also, we use the GIS data of Beijing's public facilities, such as bus stations, railway stations, park, hospital, primary school and so on. With the help of ArcGIS, GS+, Surfer and Geoda Software, we explore the spatial pattern of residential land parcels, residential land price and determinants of residential land price in Beijing. In the first place, we use the methods of Spatial Trend Analysis, Nearest Neighbor Index (NNI), Exploratory Spatial Data Analysis (ESDA) to explore the spatial pattern of residential land parcels and their price in Beijing. In the second place, we compare the spatial econometric models (SLM and SEM) with traditional OLS model to further explore the determinants of residential land price in Beijing. Based on the analysis, the main conclusions are drawn as follows. (1) The number of residential land leasing parcels is not balanced among years and ring roads. The residential land leasing parcels in the last 20 years are mainly concentrated between the fifth and the sixth ring roads in Beijing. (2) Residential land parcels are generally distributed along the main roads (such as Beijing-Shijiazhuang Expressway, Beijing-Kaifeng Expressway, Beijing-Shanghai Expressway and Beijing-Tibet Expressway) and the subway lines (such as Line 1, Line 5, Line 6, Line 15, Fangshan Line, Daxing Line and Yizhuang Line), which is more obvious in outer suburban areas. (3) Generally, there exists an inverted U-shaped curve trend, indicating that residential land price declines gradually from the city center to the city fringes as a whole, and spatial pattern of residential land price has turned from mono-centric structure to poly-centric structure. (4) Residential land price demonstrates a spatial cluster distribution pattern. There exists obvious spatial autocorrelation in residential land price and it is easy to distinguish "cold spots" from "hot spots". (5) In the model selection, we compare the spatial econometric model (SLM and SEM) with the traditional OLS model. The result shows that SLM is the best, followed by SEM, indicating that there indeed exist spatial spillover effects and spatial dependence in residential land price rather than error dependence. The residential land price is mainly affected by the surrounding residential land price, distance to bus station, distance to subway station, distance to key primary school, area of land parcel, FAR and the type of land leasing. However, in this paper, one drawback is that we fail to take macroeconomic policy factors into consideration, which may play a key role in the formation of residential land price. Also, we have not considered the subway's impact in different periods such as planning period, construction period and operation period on residential land price, which needs to be further studied.