Forest structure and ecological function serve important roles in China's Grain for Green project. This paper analyzes seven grain for green patterns in the Loess Plateau Gully and Hill Region using data analysis and modeling. Fourteen forest structure factors and two ecological function factors were selected for analysis; these factors were reduced dimensionally using factor analysis, and models of the degree of coupling and interaction between forest structure and ecological function were constructed. Based on the model, analysis of data related to forest structure, soil erosion and runoff was completed for the Grain for Green project in WuQi County, Shanxi Province, China. The results reveal various relationships exist between the different forest structural factors analyzed here. All the factors can be reduced dimensionally into four main factors identified here as F1, F2, F3, and F4. These four factors represent the tree layer, upper canopy layer mass, soils and an alternative factor, respectively. A component score coefficient matrix in the Grain for Green patterns was obtained by multiplying the factor coefficient and the corresponding basic data, and then combining the scores to calculate a component score. Each forest structural feature was expressed using the component score coefficient matrix, which shows that two types of forests, pure Robinia pseudoacacia and mixed Robinia pseudoacacia/Fructus Hippophae had the highest F1 score; pure Caragana intermedia forest had the highest F2 score; and pure Fructus Hippophae forest had the highest F3 score. The coupling degree of the Grain for Green patterns differed between different models, and the integrative indices of forest structure did not correlate well with those of ecological function. The integrated index measuring the relationship between ecological function and surface runoff is larger than the index measuring the relationship between ecological function and soil erosion. Forest component factors have the most obvious effect on ecological function, and the effect of the soil factor is the least obvious. The mass of herb layer, litter layer and some other structures near the ground play the main role in the ecological function of controlling soil erosion. The relationships of forest structure and ecological function in the grain for green project differ in the various patterns. Forest structure and ecological function were most closely related in areas with pure Fructus Hippophae, while index score belongs to the middle coupling coordinative level. Pinus tabulaeformis had the least interaction between forest structure and ecological function. Calculating the score for all types of models discussed above shows a measure of the degree of interaction between the forest structure and ecological functions can be measured with a score of 0.6, which is considered to indicate a weak interaction. Therefore, forest structure and ecological function should be improved during the application of any Grain for Green project in the Loess Plateau Gully and Hill Region.