Drought is one of the most severe natural disasters that have widespread impacts on eco-environment and agriculture. Great efforts have been made on the study of the non-stationarity of hydrometeorological processes, while few reports are available addressing non-stationarity in drought index. Therefore, in this study, we attempted to develop a non-stationary standardized precipitation evaporation index (NSPEI) based on standardized precipitation evaporation index and non-stationary theory with the aim of investigating drought conditions across China in both space and time under different emission scenarios from 2006 to 2100. The results indicated that: (1) Stations with non-stationary hydrometeorological processes are concentrated in northeast China, the Huang-Huai-Hai Plain, the Yangtze River Delta, the Tibetan Plateau, the Hengduan Mountains, and the southern Xinjiang. The NSPEI has the best fitting performance at 88% of the meteorological stations considered in this study. (2) Compared with other drought indices, SPEI tends to overestimate the intensity and duration of droughts during evaluations of the future drought changes; while NSPEI avoids the weakness of SPEI in overestimation of drought intensity. Therefore, NSPEI can better monitor the meteorological droughts in China and describe the future drought changes across the country. (3) Drought monitoring results based on NSPEI indicated an increasing drought trend in northern China, and southern China is dominated by a wetting trend under the scenario of low and high emissions. Extreme dry and wet duration and occurrence frequency showed an increasing trend in China in the future under medium and high emission scenarios.