ABSTRACT The Qinghai‐Tibetan Plateau (QTP) experienced noticeable warming and glacial retreat during the past decades. However, it is unclear how these changes affect QTP wetland distribution in the past and future. To this end, this study estimated the potential wetland distribution in the QTP under present and future climate scenarios using five machine learning methods. We further decoupled the sensitivity of wetland area to temperature, precipitation, and glacier changes based on the control experiment, and quantified the environmental niche of QTP wetland distribution. The RusBoost algorithm model has the best performance and shows that the current potential wetland area is about 1.6 × 10 5 km 2 , accounting for 6.22% of the land surface. By 2100, QTP wetlands are projected to increase by 9.6% and 77.3% relative to the current potential wetland area under the SSP1‐2.6 and SSP5‐8.5 scenarios, respectively. Climate warming and wetting are positively correlated with the future wetland areas. Each 1°C increase in the warmest season temperature can lead to a 9.0% increase in QTP wetland areas. Glacial retreat to some extent leads to wetland increase, for example, in the southeastern QTP, likely due to glacial meltwater recharge. However, wetlands will decrease due to longer glacial distances in the northeast QTP, because wetlands tend to grow within a suitable distance of 30 km to glaciers. As more current wetlands spread within the recharge range of glacier meltwater, QTP wetlands expect to increase in the near future. This research provides a valuable reference for predicting wetland changes in alpine regions in the context of global warming.