Accelerated Life Testing (ALT) is an efficient approach to obtain failure observations by subjecting the test units to stresses severer than design stresses and utilize the test data to predict reliability at normal operating conditions. ALT is usually conducted under constant-stresses which need a long time at low stress levels to yield sufficient failure data. Many stress loadings, such as ramp-stresses obtain failure times faster than constant-stresses but the accuracy of reliability predictions based on such loadings has not yet been investigated. We develop test plans under different stress applications such that the reliability prediction achieves equivalent statistical precision to that of the constant-stress. The research shows indeed there are such equivalent plans that reduce the test time, minimize the cost and result in the same accuracy of reliability predictions. © ICAQM 2011.