Abstract: Based on load uncertainty modeling, this paper presented a risk assessment framework of static voltage stability in the power system. Load uncertainty modeling consisted of three steps: Modified fuzzy cluster method（FCM）, Cholesky decomposition and multi-dimension normal distribution sampling method were adopted to get initial load profile; A probability hyper-cone load growth model was proposed based on Monte Carlo method to simulate stochastic variation of loading direction and power factor; One evaluation system combining correlation coefficient with mutual information was established to divide buses into different groups according to the similarity degree of their load curve. When load uncertainty model was established, it would be inserted into proposed Monte Carlo- continuation power flow（MCCPF） algorithm. Four risk indexed focusing respectively on bus, line and entire system were proposed and would be calculated based on MCCPF performance results, integral risk assessment framework was then founded. Taiyuan 110 k V system is studied and analyzed in terms of risk level and weakness zone recognition, adopting respectively constant power load model and composite load model, and the validity and effectiveness of proposed method is verified.