This repository includes land use and land cover maps of Mozambique for 2000, 2005, 2010 and 2016 years. The methodology is based on remote sensing methodology and include satellite image collection and compositing (annual cloud-free and shadow free Landsat images for 2000, 2005, 2010 and 2016), delineation of a large collection of training plots based on National Land Cover Classification system level 1, supervised classification using a machine learning algorithm (Random Forest) and post-processing steps. The LULCC map for 2016 show area statistics of 45.0% (35.8 Mha) of dry forest, 37.0% (29.3 Mha) of grassland and fallow, 13.7% (10.8 Mha) of cropland 2.0% (1.6 Mha) of wetlands, 1.3% (1 Mha) of other categories (rocks, sands, or bare soils), 0.3% (271,000 ha) of Mangroves, and 0.1% (673.1 ha) of urban areas. The deforestation over the 2000-2016 period is estimated to have been 207,272 ha per year. The methodology and statistics are presented in the report included in this repository. Theses maps are outputs from the study "An Analysis of Land Use Changes and Land Degradation in Mozambique" conducted by Nitidae and CIRAD in the LAUREL project.