Using satellite and field data to investigate soil moisture is a developing process challenged by the landscape complexity of most terrestrial ecosystems. Landscapes are complex systems formed by spatial units (fragments) in which biophysical factors fluctuate according to soil and landuse characteristics. This research uses a multidisciplinary approach to explore the relationships between fragments and process in complex landscapes. I investigate at the Little River Watershed (LRW) Georgia US, how the spatial complexity of the landscape caused by different combinations of soil types and land uses influence the spatial extension of soil moisture and ground temperature within areas equivalents to the pixel sizes of four environmental satellites. The purpose was to define landscape-based criteria to use satellite data in direct and indirect assessment of soil moisture and to define the spatial extend for which single point readings can be used to estimate soil moisture conditions under complex landscapes. The methods include the creation of a GIS database for soil and land cover at <10m spatial resolution, distance and area spatial analysis for field point data and hydrological modeling for five different soil types. Supported by rigorous statistical analysis such as analysis of variance, t-test, descriptive statistics, correlation analysis, time stability analysis, and generalize linear model, the results include: 1. Criteria to define the appropriate pixel sizes of a satellite instrument to study SM and ground temperature at the LRW. 2. Understanding effects of landscape and local complexity and their weight in point readings. 3. The assessment of SM responds by different soils and land cover combinations. Overall, this research contribute to satellite study of the water cycle by proposing a landscape ecology–hydrology approach to the problems of point reading interpolation and remote sensing of SM within complex landscapes.