Taking Xinjiang, a typical inland arid region of China, as the study area, this study applied the empirical quantile mapping method to correct the biases of precipitation and temperature outputs from 30 CMIP6 global climate models, comprehensively evaluated their applicability in Xinjiang, and selected an optimal multi-model ensemble to predict future precipitation and temperature under different emission scenarios. The results show that Xinjiang will experience a significant warming and humidification trend in the future. Compared to the baseline period of 1995-2014, by 2100 under the SSP1-2.6, SSP2-4.5, SSP3-7.0, and SSP5-8.5 scenarios, annual precipitation is projected to increase by 15%(21mm), 23% (33mm), 25% (36mm), and 48% (69mm), respectively, and annual mean temperature is projected to rise by 2.3℃, 3.6℃, 5.3℃, and 6.7℃. The largest precipitation changes are expected in Hami City, Bayingolin Mongolian Autonomous Prefecture, and the southern part of Hotan Prefecture, and the increased precipitation may help alleviate ecological water replenishment pressure in the lower reaches of the Tarim River. High temperature increase zones include the border areas of Tacheng Prefecture, Altay Prefecture, Urumqi City, and Changji Hui Autonomous Prefecture, as well as most areas of Turpan City, Bayingolin Mongolian Autonomous Prefecture, Hotan Prefecture, and Kashgar Prefecture, and the maximum temperature rise could reach 7.6℃ in the far future (20762100). The future temperature rise will lead to exacerbated desertification in Turpan City, Bayingolin Mongolian Autonomous Prefecture, and Hotan Prefecture, and intensify soil salinization and alkalization and farmland degradation in some areas.