Overall survival (OS) is recognized as the gold standard for clinical benefit in oncology clinical trials. However, the formal demonstration of the underlying survival benefit of a treatment can be diluted or confounded by effective later treatment. Although oncology trials have the objective of comparing the experimental arm with the control arm for OS, statistical analyses of OS by Intention-to-treatment (ITT) treatment groups based on conventional approaches are biased towards the null because of the crossover. This has been a great challenge in oncology drug development and regulatory approval. In this roundtable luncheon, we hope to stimulate discussions about the possible approaches in designing trials and estimating the underlying treatment effect of OS when crossover is present.