The classical theories of reliability parameter sensitivity analysis based on probability theory and fuzzy mathematics cannot directly solve the uncertainty of subjective,random and fuzzy variables.To solve the problem,the uncertainty theory is introduced,and uncertain variables are utilized to uniformly represent the subjective,random and fuzzy variables,so as to satisfy the duality of randomness and subadditivity of fuzziness.Moreover,new formulas of reliability parameter sensitivity analysis are derived based on Cornell uncertain reliability index.Numerical examples have demonstrated the effectiveness of the proposed method, which can provide a reference for reliability optimization.