Optimum simple step-stress accelerated life testing plans have practical limitations, They highly depend upon the assumption of a linear relationship between stress and time-to-failure and use only two extreme stresses that can cause irrelevant failure modes, Thus, 3-step plans are preferable, In this paper we derive the optimum quadratic plan and propose and evaluate a 3-step test plan, the compound linear, as an alternative to the optimum simple step-stress plan and optimum quadratic plan. The compound linear plan is better than two compromise plans proposed for constant stress testing and nearly as good as a proposed optimum compromise plan, the advantage being its simplicity, A small sample-size study shows that the efficiency of the test plan is not appreciably affected by the sample size, so that the asymptotic results are a good indication of the performance in sample sizes encountered in practice.