Land use change is certainly the most important factor that affects the conservation of natural ecosystems, resulting the conversion of natural lands such as forests and pastures into agricultural, industrial and urban areas. Despite numerous studies investigating landscape patterns due to land use change, the driving forces of landscape change has been less studied in Iran. In this study, Artificial Neural Network (ANN) method was used to investigate the process of landscape change using ten variables including slope, distance from built-up areas, water bodies, road, forest edge, rangland and agriculture, number of forest classes and elevation. Aspect and distance from water bodies variables were removed based on the Cramer’s V statistic. Using transition potential maps, land cover distribution patterns for the year 2032 were created. Also, the relative effects of the 10 predictor variables were evaluated through the sensitivity of the model by forcing a single independent variable to be constant. Distance from rangeland and distance from built-up areas were the most influential variables on land use change. Kappa coefficient was used to assess the accuracy of the modeling approach. Kappa value for ANN was 0.82. We also used landscape analysis to compare modeling results through landscape change process. The general pattern of land use change in Gharesoo Watershed showed that the landscape change process related to human (built-up areas and agricultural lands) was in the form of creation and aggregation and the category of change for natural uses (rangeland and forestland) was in the form of loss and fragmentation. Introducing "Landscape Change Process" approach in this study provides a comprehensive understanding of changes in the landscape configuration for each land use class by simplifying the analysis.