Global climate change has intensified the frequency and severity of compound dry-hot events (CDHEs), posing more severe impacts on terrestrial ecosystems than individual extremes, particularly in Northwest China (NWC). However, a comprehensive probabilistic assessment of vegetation vulnerability under CDHEs, particularly, the identification of triggering thresholds for vegetation loss remains limited. This study employed standardised indices of vegetation (NDVI), drought (SPI) and hot conditions (STI) from 1982 to 2018 to develop a Vine copula framework for assessing vegetation vulnerability under CDHEs and identifying the corresponding triggering thresholds for vegetation loss. The results demonstrated that the standardised compound event indicator (SCEI) in NWC indicated a decreasing trend (-0.12 decade(-1)), reflecting the intensification of CDHEs. This intensification of CDHEs has led to greater ecosystem vulnerability in NWC. Under Severe Dry-Severe Hot conditions (S4: SPI < -1.3 and STI > 1.3), the average probabilities of vegetation loss below the 50th, 30th, 20th and 10th percentiles were 51.2%, 42.3%, 36.2% and 27.2%, respectively. Different vegetation types (croplands, forests and grasslands) exhibited distinct vulnerability patterns, with grasslands being the most sensitive and forests the least. The average SPI/STI thresholds corresponding to vegetation loss below the 50th percentile (mild) and 10th percentile (extreme) were -0.71/0.72 and -1.08/1.35, respectively. It could provide a novel framework for assessing vegetation vulnerability under CDHEs, while simultaneously offering a profound comprehension of ecosystems' response mechanisms to CDHEs.