The need to control the potential for error and to report claims of scientific findings in a principled and parsimonious way has been recognized for decades and has spawned the development of a substantial corpus of statistical methodology. However, the discussion of the issues was confined to multiple testing, remained mostly in the regulatory realm, and never attracted such broad and intense interest as it did in recent years. This interest is only partially reflected by in the multitude of commission reports, position papers, and organized discussions in scientific forums. Indeed, the growing concern about the reproducibility of scientific findings and the multiplicity of reported inferences has led to concrete, sometimes sweeping changes across scientific fields. In particular, the impact of these developments is beginning to be felt strongly in biomedical research. In this presentation we will take stock of important changes in reporting practices, highlight intended and unintended consequences, and point to areas where the scrutiny needs to be extended and where further methodologic development is needed.