Statistical inference is considered on constant-stress accelerated life test when the failure data are progressively Type-II censored. Under the assumption that the Weibull shape parameter is nonconstant and both Weibull parameters follow log-linear life-stress model with stress, the unknown coefficient parameters are estimated by using methods of maximum likelihood and expectation–maximization based estimations as well as approximation maximum likelihood estimation. The confidence intervals of unknown parameters are also constructed based on asymptotic theory and bootstrap technique. Simulation study and a real data example are presented for illustrative purpose.