In this paper, climatic and environmental changes were reconstructed since 1129A.D.based on the Malan ice core from Hol Xil, the northern Tibetan Plateau. The record of δ 18 O in the Malan ice core indicated that the warm-season air temperature variations displayed a general increase trend, the 20th-century warming was within the range of natural climate variability, and the warmest century was the 17th century while the warmest decade was the 1610s, over the entire study period. The "Medieval Warm Epoch" and "Little Ice Age" were also reflected by the ice core record. The dust ratio in the Malan ice core is a good proxy for dust event frequency. The 870-year record of the dust ratio showed that dust events occurred much frequently in the 19th century. Comparing the variations of δ 18 O and the dust ratio, it is found that there was a strong negative correlation between them on the time scales of 10 1 -10 2 years. By analyses of all the climatic records of ice cores and tree rings from the northern Tibetan Plateau, it was revealed that dust events were more frequent in the cold and dry periods than in the warm and wet periods.Keywords: the Tibetan Plateau, the Malan ice core, climate changes, dust event frequency.Conventional wisdom on the climate of the last millennium is very simple, a "Medieval Warm Epoch" (MWE), a "Little Ice Age" (LIA) and then globally extensive warming. With the accumulations of the paleoclimatic data from different regions of the globe, this view, based almost extensively on evidence from western Europe and the North Atlantic region [1,2] , has been reassessed [3][4][5][6][7] . Moreover, it was revealed that air temperatures in some areas went down in the recent decades [8][9][10] , even though the most areas of the globe were warming. These suggest that climate changes have temporal and spatial differences. Recently, several studies have reconstructed the variations of hemispherical or global air temperatures over the last millennium [11][12][13][14][15] . These reconstructions for the early time period might be possibly biased towards the variations of air temperatures in the regions with long data sets, because the number of the samples of the data sets longer than 400 years was small. There also exist some differences among these reconstructions, which might be caused by the fact that the different methods were used and/or the data from different regions were 1080 Science in China Series D: Earth Sciences selected. However, to reveal the nature of air temperature variations in the hemispherical and/or global scales and the temporal and spatial differences of the air temperature variations in different regions is the basis for understanding the inner and exterior causes of climate change and predicting the future climate change. In order to better resolve these issues, much more new paleoclimatic records with long period and high resolution need to be gained from different regions, especially from the regions where paleoclimatic data are lack. In this paper, we will reconstruct a recor...