DescriptionAccelerated life testing of manufactured units is performed in order to speed up testing through either reducing the time required for testing or establishing a predetermined number of failures to stop the test. In this research we develop two cost models for a Type-II censored testing for Weibull distribution life time units. We determine the optimal sample size on test which minimizes the expected total cost of performing the life testing when the parameters of the Weibull distribution are either fixed or unknown. Numerical examples are included to illustrate the cost models based on real data applications.