The bootstrap weighted-norm method was proposed to evaluate the optimum confidence interval of reliability with three-parameter Weibull distribution.Based on the difference between the empirical values and the theoretical values of reliability

an information vector of the optimum parameters of the Weibull distribution was structured in the light of six criterions of the minimum weighted-norm.By bootstrap resampling the information vector of the optimum parameters

a generated information vector was obtained to simulate the probability density function of the parameters.The estimated truth-value and its confidence interval of the parameters were solved and the estimated truth-value function and its optimum confidence interval function of the reliability could accordingly be established under the given confidence level.The case studies of the rolling bearing performance life

the helicopter component failure and the specimen fatigue life indicate that the estimated truth-value function is in accord with empirical values of reliability by using bootstrap weighted-norm method

and the optimum confidence interval function is able to describe true state of the test results.