Compared with the plain area, the town's social-ecosystem in shallow mountain area is more vulnerable to the constraints of topographic environment and other factors, and needs to be weighed in the development and protection of land space. Using the SERV (Spatially Explicit Resilience Vulnerability) model, this paper constructs an evaluation system from three dimensions of exposure, sensitivity and adaptability to evaluate the social ecosystem vulnerability of 12 towns in Pinggu shallow mountain area, and uses OWA (Ordered Weighted Averaging) to simulate the vulnerability scenarios under various decision risks. The results show that:(1) from 2007 to 2017, the overall vulnerability of Pinggu shallow mountain area presented a moderate vulnerability level, and the local vulnerability level showed an upward trend. In the past 10 years, the proportion of vulnerable towns above medium increased from 73.09% to 80.74%. Spatially, it presents a pattern of high in the southeast and low in the northwest. (2) Towns at a high vulnerability level should be controlled to increase the area of soil and water conservation forests; At the same time, we should pay attention to the high exposure risk sources of low vulnerable towns, carry out ecological restoration in time, and narrow the vulnerability difference between the north and the south. (3) By setting different decision-making risk coefficients, the future scenarios predict the vulnerability differences of regional systems under different development orientations. The government gives priority to the management of the ecological environment under the sustainable development scenario and promotes the construction of green infrastructure under the economic development scenario. The evaluation results can meet the requirements of guiding regional development practice under different decision-making ideas.