The quality of research depends on the rigor with which researchers conduct studies and control for potential biases. Recent publications and meetings have focused on the observation that published research results are often not reproducible. Although there are many possible causes for this problem, poor experimental design and inadequate reporting of research methodologies appear to be significant contributing factors. Evidence for this deduction will be presented, as well as actions taken by the US National Institutes of Health (NIH) and by publishers to address the issues.