Expansion of the spatial coverage of pollen data is essential to improve understanding of Holocene climate variation. To address this, we collected 25 surface and 72 fossil samples from Butuo Lake in the eastern Qinghai–Tibetan Plateau, China. We reconstructed the vegetation history of the region through detrended correspondence analysis (DCA) of the pollen data. Based on the results of this analysis, we divided the samples into five pollen zones (alpine sparse cushion vegetation, alpine scrub, alpine meadow, montane scrub meadow, and montane coniferous forest) corresponding to the major vegetation types. The observed temporal changes in vegetation (as indicated by the DCA of surface and fossil pollen spectra results, ratios of Artemisia to Cyperaceae (A/Cy), the sum of percentages of dryness indicators (SDI), and percentages of main pollen types) and modern pollen–climate transfer function (developed using the weighted averaging partial least squares regression method (WAPLS)) yield a sensitive record of Holocene monsoonal climate change in the area. During 11,140–8700, 8000–6000, and 5600–1780 cal. yr BP, the climate was wet and the vegetation was dominated by alpine meadow, indicating the occurrence of a strong southwest Asian monsoon that spanned almost the entire Holocene. Notably, two major cold and drought episodes are detected at 8700–8000 and 6000–5600 cal. yr BP, with vegetation dominated by alpine steppe, suggesting that the southwest Asian monsoon was extremely weak during these periods. The data will improve understanding of long-term variations of the southwest Asian monsoon in the region.