Non-inferiority trials are a valuable type of clinical trial to evaluate new treatments that may have different benefits. These trials are increasingly used, but they have methodological problems in their design and analysis. These issues must be considered in promoting research and publications of the highest quality. The commentary lists key points to consider. They involve the selection of a non-inferiority margin (M or Δ), which is established before the trial and determines the degree of inferiority of the test treatment to the control. If the confidence interval for the difference excludes that the non-inferiority is greater than the margin, the treatment is declared non-inferiority. Analysis should use ITT (intention-to-treat) and PP (per protocol) sets equally important to lead to similar conclusions.