Accelerated lifetime data can be analyzed either by parametric or nonparametric methods. This thesis reviews the currently available nonparametric approaches to accelerated life testing(ALT). Nonparametric ALT models are classified into the stochastic ordering models and the time acceleration function model, and advantages and limitations of each model are discussed. For the time acceleration function model, a test procedure is developed to determine whether the time acceleration function is a linear or a power type.