Principal stratum strategy is used to address intercurrent events when the treatment effect is of interest on the subpopulation defined by a post-randomization variable, known as principal stratum. Statistical analysis for principal stratum estimand involves statistical modeling and a series of assumptions because it is impossible to directly summarize the treatment outcomes over a principal stratum. In this chapter, we will go through a few common assumptions for principal stratum strategy and the statistical analyses based on them. Sensitivity analyses are also briefly reviewed. We will also discuss the special consideration for principal stratum analysis when outcomes are time-to-event. Lastly, we listed some cases of principal stratum estimand in this book.