DSMBs play an important role in protecting the safety of trial participants while preserving the scientific integrity of the trial. This requires that they be able to balance potential safety concerns with the potential for the trial to demonstrate benefit. Thus, DSMBs must receive summaries of evolving results that are maximally informative and easy digest. Trials using sophisticated computational algorithms for making adaptations, incorporate an additional layer of complexity. To the extent that the DSMB is directly involved with the adaptations, these should be in addition, rather than in place of their normal data monitoring activities. Furthermore, DSMB decisions must involve the totality of evidence and require the flexibility to make recommendations independent of the adaptive procedures. This requires that the adaptive procedures accommodate for some degree of DSMB discretion without jeopardizing the operating characteristics of the trial procedures. This talk will discuss some the competing interests that DSMBs face and make recommendations regarding how these interests can be accommodated.