Abstract: Abstract: Under rapid urbanization, analyzing the spatial characteristics and pattern types are fundamental tasks of revealing the mechanism of formation, growth and transition of rural settlements, which are essential to make reasonable decisions for rural settlements consolidation activities. The existing researches mainly considered rural settlements size or configuration characteristics at national or provincial scale, which inevitably hided various problems in small regions and failed to provide explicit characteristics of current size, shape and location required with readjustment of rural settlements. Thus this paper firstly defined rural settlements pattern as the composition of various residential land parcels with number, size, shape, location and organization relationship characteristics. Based on the definition of spatial characteristics of rural settlements pattern and the main compositions, the indicators system and characteristic index model with geometric mean method for evaluating the spatial characteristics of rural settlements pattern were built from size, configuration and location three dimensions. Then a "3-3-2"combination matrix was designed to calculate types of rural settlements spatial pattern by dividing size characteristic index into 3 grades (i.e., small size, medium size and large size), configuration characteristic index into 3 grades (i.e., irregular, moderate irregular, regular), and location characteristic index into 2 grades as advantageous condition and disadvantageous condition. Thus in theoretical rural settlements pattern was concluded 18 types (i.e., 3×3×2=18) of characteristics discussed above. Finally, the paper took Pinggu district in Beijing as the study area, analyzed the spatial characteristics and types of rural settlements pattern, and recommended optimal directions for rural settlements development in the future.Results showed that rural settlements pattern had multi-dimensional characteristics, and the quantitatively analysis from spatial size, configuration and location reflected the current state of distributation. The combination matrix method effectively overcame the limitations that comprehensive evaluation method hided the variation of pattern's dimensional characteristics. The study showed that rural settlements in Pinggu existed great regional variations, which in size were larger in plain region, came next in semi-mountainous region, and were smaller in mountainous region. Additionally, some rural settlements in configuration confronted irregular and scattered problems, though in location most rural settlements presented the tendency of nearing regions with low altitude, roads, city and towns. Pinggu's rural settlements patterns were divided into 10 types, and mainly included "medium size-moderate regular configuration-advantageous location" and "medium size-regular configuration- advantageous location". The study proved that rural settlements patterns were both influenced by geographical factors and socio-economic factors, but they acted on different characteristics of the pattern. The geomorphic condition basically controlled the size characteristic of rural settlements, while the infrastructures like roads and the adjacent central towns (city) further influenced on their spatial configuration and location. Rural settlements consolidation was a complex program that needed to optimize the hierarchy system from macro perspective, and that also needed to diagnose detail problems about size, configuration and location characteristics at small scale. The study indicated that types of rural settlements pattern indicated by the combination matrix method efficiently reflected multi-dimensional information, which immediately provided guiding directions for rural settlements consolidation and spatial optimization practices.