Based on the structure and function of water resources system, a water resources vulnerability evaluation system is constructed, which consists of three criterion layers, namely, water yield, water quality and disaster, three constraint layers, which are natural factors, human factors and bearing factors, and 27 indices. The weight are assigned using the combination of entropy weight method and analytic hierarchy process, and the water resources vulnerability of 10 prefecture-level cities in Shaanxi Province from 2007 to 2017 is evaluated and categorized using the fuzzy comprehensive evaluation method from spatial and temporal perspectives. In the meantime, the obstacle degree model is applied to explore the obstacle factors of water resources vulnerability. The results show that the water resources vulnerability of Xianyang and Weinan is categorized as Grade Ⅴ; that of Xi’an, Baoji, Tongchuan and Yulin is Grade Ⅱ; and that of Yan’an, Hanzhong, Ankang and Shangluo is Grade Ⅰ. Over the 11 years, the water resources vulnerability of Xianyang, Weinan, Ankang and Shangluo showed a downward trend, whereas that of Xi’an, Tongchuan, Yan’an and Hanzhong showed an increasing trend, and that of Baoji and Yulin remained unchanged. Spatially, it showed the distribution pattern of “low in the north and south regions, high in the middle regions”. The main obstacle factors affecting water resources were change rate of annual precipitation, water consumption of 10 thousand RMB GDP, disaster ratio of drought and flood area, wastewater discharge of 10 thousand RMB GDP, and development and utilization rate of groundwater resources. The research shows that the water resources vulnerability varies widely across various regions in Shaanxi Province, with the coexistence of the tendency of increase and decrease, indicating that individualized management and regulation measures that suit local circumstances should be carried out.