Abstract
In this paper, optimum test plans of step-stress partially accelerated life tests are developed using failure-censored data from Burr type-XII distribution. The maximum likelihood approach is applied to obtain the estimates of the acceleration factor and the parameters of the distribution. The minimization of the generalized asymptotic variance of the maximum likelihood estimators of the model parameters is used as an optimality criterion to develop optimum plans of the step-stress partially accelerated life tests. For illustrative purposes, simulation studies are presented.