The development of human society is associated with the expansion of land use over natural areas. Vegetation can be obliterated by urban expansion and agriculture. Therefore, the aim of this study is to identify, using temporal analysis and landscape metric, the natural vegetation remaining in Mogi-Mirim/SP. Images from Landsat5-TM system sensor were acquired and an RGB composite color, using Envi 5.3 software, were performed. Further processes included segmentation and classification, to separate the different objects, using e-cognition software. The classification process is based on the Object-Based Image Analysis (OBIA) paradigm. Landscape metric were calculated, by the extension Vector-based Landscape Analysis Tools (V-Late) 2.0 beta in Arcgis 10.3 software. The metrics analyzed were area, form, proximity and nearest neighbor. The results show that natural vegetation in 2011have been obliterate due urban, agriculture expansion when compare with 1990. Besides the major number of fragments in 2011, because the confusion between natural vegetation and Pinus, in this year the landscape were more fragmented. As a consequence of fragmentation, there´s changes in biological diversity and animal fluxes. These remote sensing techniques are efficient to identify remaining forest fragments and evaluate the state of landscape with landscape metrics, and can be effective to landscape monitoring. Future works recommendations include try different segmentation algorithms