The paper is to improve the famous contingent claims analysis (CCA) approach. It first demonstrates that the core assumption of Merton-CCA model is unrealistic and is not apt to measure risk in a realistic，dynamic economic environment. Then，a probability model with ambiguity，which incorporates model uncertainty，volatility uncertainty and mean uncertainty，is proven to be more effective for measuring and managing risk in the real business world. Based on the recent progress in stochastic analysis and calculus，the paper proposes a G-CCA model，a more risk sensitive，thus a more prudential risk measurement model，especially in emerging financial markets. Moreover，adding to all the valuable advantages of the Merton-CCA model，the new model extends the risks covered by traditional CCA models by taking into account risks arising from the uncertainty of expected asset returns.