More and more clinical trials are multi-regional. Regional heterogeneity is not rare but often extremely difficult to handle when it appears in practice. On one hand, many argue that regions should be considered as subgroups and thus the issues should be handled like subgroup analysis issues. On the other hand, many articles are published to investigate the problems of consistency or inconsistency between regional results and how to handle them. Should regions be defined based on geographical information or something else? What is the major difference between multi-regional issues and multi-center issues? How should the potential inconsistency be considered in analysis and design stages? This presentation will focus on discussing these issues and suggest steps forward.