Land use systems are a fundamental part of the Earth's surface. Change of land use has significant impacts on climate, biodiversity, hydrological cycles, biogeochemical processes and human society. The Mau forest complex in Kenya is a montane forest ecosystem where significant land use changes have occurred: over the last few decades a quarter of the forest cover has been lost. The complex is one of the largest closedcanopy montane forests in Eastern Africa and is part of the global biodiversity hotspots; ecosystems that are exceptional rich in biodiversity and where abrupt and significant environmental degradation takes place. In addition, the forest fulfils a crucial socio-economic function because several millions of Kenyans rely on the products and services it provides. Currently there is no comprehensive overview of the driving factors of land use change in the Mau forest complex. Knowledge of these drivers is important for policy making and modelling future processes. To address this gap, this study analysed the land use changes and its drivers in the Mau forest complex in the period 1973-2013. Remote sensing and GIS techniques combined with multiple logistic regression modelling were used to identify the dynamics and drivers of land use change. This study shows that the main land use changes in the Mau forest complex in the period 1973-2013 were loss of forest and rangeland, while smallholder agriculture extended. More precisely, the largest land use change in the study area was a conversion from forest to smallholder agriculture. Hence, smallholder agriculture can be considered the most important proximate driver of deforestation in the Mau forest complex in every time period analysed. Based on the accuracy assessment and land use change dynamics analysis two land use change models were fitted: forest conversion and smallholder agricultural expansion. The regression analysis showed that biophysical and socio-economic factors were significant driving forces in both models. Drivers such as aspect East, curvature, the topographical wetness index, population density, distance to towns and distance to roads increased the odds of forest conversion, and in particular the distance variables became more important in more recently periods (1994-2003 and 2003-2013). In the agricultural expansion model, biophysical factors had mainly an influence in the second period (1984-1994), while the socio-economic underlying drivers were more important in the third and fourth periods (1994-2003 and 2003-2013). In the overall period 1973-2013, both the forest conversion and agricultural expansion model showed that a growth in population density increased the chance of land use change. In conclusion, this research demonstrates that land use change and its drivers show different spatial-temporal trends. The models revealed an increasing importance of socio-economic variables which means there is a need for better understanding the socio-economic aspects behind land use change. Therefore, future research and policies should be time and space specific and focus more on socio-economic drivers of land use change.