Humans are currently facing serious environmental challenges at global and local levels. Sustainable development has become equally important with national security in recent years. The sustainable ecological carrying capacity of an ecosystem is best indicated by the relation between ecological carrying capacity and ecological load. The present research aimed to develop a forecasting system for measuring sustainability of ecosystem in Wulingyuan Scenic Spot using an early warning of ecological carrying capacity. For this, the evaluation indices system of ecological carrying capacity in Wulingyuan Scenic Spot was established using principal components analysis (PCA) and analytic hierarchy process (AHP). On this basis, 17 indices including the resources, environment and social-economy, which were separated from the resource carrying capacity, environment carrying capacity, ecological elasticity and ecological pressure-bearing, were chosen as the early-warning indices inputs. Afterwards, the whole Wulingyuan Scenic Spot was divided into several early warning districts with different sustainability of ecosystems according to the ecological load-bearing index calculated by the State Space Approach (SSA). Finally, based on the samples tests, the BP neural network model (BPNNM) was established, and the forecasting system for ecological carrying capacity in Wulingyuan Scenic Spot was developed. The results showed that the whole Wulingyuan Scenic Spot was over the carrying capacity in 2000, which was mainly due to the urbanization of core areas in it. The results further revealed that Xiehe Village was ever over the carrying capacity because of the unreasonable exploitation and utilities of the local resources. The forecasting system of ecological carrying capacity, carried out by the BPNNM, might be useful in forecasting sustainability of ecosystems. Therefore, it can promptly reflect the regulation consequences of ecosystems sustainability and serve as scientific basis for the regional sustainable development.