In this chapter, we explore statistical analysis of binary endpoints under the estimand framework. While the consideration for addressing intercurrent events and the principle for missing data imputation for binary endpoints are similar to the continuous variables, the statistical methods and the summary statistics for binary endpoints need further discussion. We briefly introduce the common statistical methods for binary endpoints and review three commonly used summary measures: risk difference, relative risk, and odds ratio. We delve into the similarities and differences among these summary measures from the logic-respecting perspective and provide recommendations. Additionally, we illustrate the tipping point analysis, which is commonly used for binary endpoints in the sensitivity analysis. Finally, we present and compare the original analysis method and the analysis method under the estimand framework from a real case study.