Many pivotal study protocols propose sample size re-estimation (SSR) during the midcourse of a clinical trial. Two strong motivations are usually cited in the literature: first, the original sample size is often estimated based on insufficient or inaccurate information; second, the sample size estimate based on a clinically meaningful treatment effect is often overpowered, which implies an unnecessary large sample size and a waste of resources. There are mainly two types of SSR approaches. Some proposed adjustment upward only by updating a nuisance parameter to ensure sufficient power. Others have proposed SSR based on the interim finding of the treatment effect size rather than the nuisance parameters. In this round table discussion, the pro and cons of SSR approaches and the challenges to FDA reviewers will be discussed.