Snow water equivalent (SWE) is a critical parameter for characterizing snowpack which has a direct influence on the hydrological cycle, especially over high terrain. In this study, SWE from 18 Coupled Model Intercomparison Project Phase 5 (CMIP5) model simulations is validated against the CanSISE SWE. The model simulations under RCP8.5 and RCP4.5 are employed to investigate projected changes in spring/winter SWE over the Tibetan Plateau (TP) under global warming of 1.5°C and 2°C. Most CMIP5 models overestimate the CanSISE SWE. A decrease in mean spring/winter SWE for both RCPs over the most regions of the TP is predicted in the future, with most significant reductions over the western TP, consistent with pronounced warming in that region. This is supported by strong positive correlations between SWE and mean temperature in the future in both seasons. Compared with the pre-industrial period, spring/winter SWE over the TP under global warming of 1.5°C and 2°C will reduce significantly, at faster rates than over China as a whole and the Northern Hemisphere. SWE changes over the TP do not show a simple elevation dependency under global warming of 1.5°C and 2°C, with maximum changes in the elevation band of 4000-4500m. Moreover, there are also strong positive correlations between projected SWE and historical mean SWE, indicating that the initial conditions of SWE are an important parameter of future SWE under specific global warming scenarios.