With the help of remote sensing images (Landsat TM in 1990s and 2000s, Spot4 in 2006), DEM of 90m spatial resolution, 1∶50000 topographic maps, socio-economic statistical data and other relevant information, using ENVI 4.7, ArcInfo 9.2 and spatial statistical analysis software (Geoda 095i) and some other research methods such as remote sensing information extraction, spatial analysis method, land-use dynamic changes model and theory, we carry out the research on construction land-use change in Beijing-Tianjin-Hebei Metropolis Circle. Firstly, interpreting remote sensing images and analyzing the changes of urban and rural construction land-use, including the quantity and speed of land-use change, spatial pattern and spatial autocorrelation of land-use growth, the study shows that the urban and rural construction land is expanding continually from 1990 to 2006 with a 2.81% annual increasing rate, the total area of construction land is 11001.81 km², 13786.14 km², 15940.55 km² in 1990, 2000 and 2006 respectively, the relative rate of change in Tangshan, Baoding and Shijiazhuang are larger than the average rate of the total area in the period from 2000 to 2006. Secondly, discussing the spatial variation of urban construction land by spatial buffer analysis methods, the result shows that the uneven development of spatial expansion including centralized development and ribbon-shaped development is significant, that is, 70% of the new distribution of urban land is within 40km from the town centre, and 80% of the new construction sites focus on the highway along the 20km range. Comparing the spatial distribution change during the two periods, i.e., new urban land from1990 to 2000 and from 2000 to 2006, we found that the new added urban land presents an expansion trend to outskirts of the city and the highway. Finally, in order to explore spatial aggregation of new added rural-urban construction land further, Global Moran’I, Moran scatter plot and LISA spatial aggregation map with statistical unit of the 123 counties in Beijing-Tianjin-Hebei Metropolis Circle were constructed. The results show that the spatial distribution of the rural-urban construction land-use growth is not entirely random, but is spatial clustering with the similar spatial value. Moran's I of the additional urban and rural construction land during the two periods were 0.2645 and 0.4067. This study will be helpful for land use planning and ecological researches for Beijing-Tianjin-Hebei Metropolis Circle.