Given the important role of lake ecosystems in social and economic development, and the current severe environmental degradation in China, a systematic diagnosis of the ecological security of lakes is essential for sustainable development. A Driving-force, Pressure, Status, Impact, and Risk (DPSIR) model, combined with data screening for lake ecological security assessment was developed to overcome the disadvantages of data selection in existing assessment methods. Correlation and principal component analysis were used to select independent and representative data. The DPSIR model was then applied to evaluate the ecological security of Dianchi Lake in China during 1988-2007 using an ecological security index. The results revealed a V-shaped trend. The application of the DPSIR model with data screening provided useful information regarding the status of the lake's ecosystem, while ensuring information efficiency and eliminating multicollinearity. The modeling approach described here is practical and operationally efficient, and provides an attractive alternative approach to assess the ecological security of lakes.