Urban rail transit possesses significant impacts on land use change and urban development. This study applies Future Land Use Simulation Model (FLUS) to reproduce land use changefrom 2000 to 2010 in the Mid-Zhejiang urban agglomeration based on GlobeLand30 datasets. The simulation results in 2010 show that the FLUS model can reproduced a realistic land use pattern with an overall accuracy of 89.74% and FOM 29.86%. A Markov chain is then used to predict the total land demand in 2030 for predicting future land use scenarios. We design two scenarios: the scenario of business-as-usual (BAU-scenario) and the scenario based on planned urban rail transit sites (RTS-scenario). Within 5 km from the urban rail transit, the RTS-scenario yields a significant effect on built-up areas with an increasing expansion intensity, where the newly built-up areas are allocated in the suburb sand are greater than that produced by BAU-scenario by 45.25 km ².The newly built-up cells mainly occupy high-quality farmland. The farmland transformed to built-up area is higher in RTS-scenario than in BAU-scenario by 33.34 km ².We categorize the built-up expansion intensity (BUI) into five levels: lowest, low, medium, high and highest. The BUI for RTS-scenario is higher than that for BAU-scenario because the former’s proportion of expansion intensity above the lowest level is 3.70% greater than of latter. Spatial patterns for forest, grassland and water are similar between both scenarios. This study not only indicates that FLUS can be used to capture land use change and predict future scenarios, but also helps to examine the effects of urban rail transit site plansin the Mid-Zhejiang urban agglomeration.