Abstract: Spatial and temporal dynamics of land use/land cover changes were quantified using TM/ETM+images. Time series were selected for forest cover change evaluate in the North of Iran in 1989-2000. In this study, we used a supervise classification algorithm and five techniques based on thresholding involved radiance/reflectance band differencing, NDVI differencing, tasseled cap, change vector differencing and NDVI ratio. Between five change detection approaches, NDVI differencing approach was the best method for changes detecting occurred in the study area. According to measurements from satellite images, 4843.42 ha were detected in this area in 1989-2000. Man-made expansion in the forest North of Iran has been largely derived by population growth and economic development. Land use maps produced will contribute to both the development of sustainable management land use planning decisions and also for forecasting possible future changes in growth patterns. There is a merit to each of the several land use change detection methods studied and appears to be no single best method in which to perform change analysis. The resulting different spectral response of types of disturbances can be used to classify and forecast natural and man made disturbances and artificial neural network or knowledge-based expert offer further opportunities.