In the FDA’s presentations, we will introduce the concept of collective evidence in reviews of medical products. We propose to control the error rate of wrongly approving an ineffective product by using evidence collectively from multiple dosing regimens and multiple endpoints across multiple studies, and to identify the optimal doses by good evidence and understanding of dose response relationship as well as product safety profile, rather than multiplicity adjustments. We will also discuss the level of consistency in information across studies by proposing controlling error rate strongly and weakly at different levels. Cases from regulatory reviews are presented.