This study presented an economic model for estimating the land use structure changes based on Landsat Thematic Mapper (TM) imagery. The Lorenz Curve and Gini coefficient, derived from Economics, were introduced as tools for investigating and quantifying the land use structure changes from 1987 to 2007 in Xiamen, China. The results showed that the spatial Lorenz curve had the phenomenon of “three near and three far” and Gini coefficient had “three increased and three decreased”, which demonstrated significantly differed profile between various land uses on temporal-spatial change in this region. The spatial Lorenz curves of cultivated land, forest and bare land were near the perfect equality line in 1987 and moved away a little bit after twenty years. However, for built-up land, water, and wetland, they reflected opposite characteristics and approached the perfect equality line with a substantial degree, which meant the distribution of built-up land, water and wetland tended to be more uniform, while the cultivated land, forest and bare land tended to be slightly uneven distributed during the past two decades. While the spatial Lorenz curve presented a graphical view of the cumulative exposure of land use area, the Gini coefficient distilled this data even further, provided a single-parameter measure of clustering. A maximum increase of 0.031 and a maximum decrease of 0.135 in Gini coefficients showed that the major process of land use distribution had becoming more symmetric. However, the Gini coefficient cannot fully describe the features of the spatial Lorenz curve to discover the differences in spatial distribution of land use structure. So, it'll be beneficial to pay more attention to the research of quantitative indicators between the geometry features of the Lorenz curve and the land use structure in future studies.