AbstractThe number of studies designed specifically to demonstrate therapeutic equivalence or alternatively non‐inferiority of pharmaceutical treatments has increased dramatically in recent years, during which time awareness of the methodological issues has increased. Regulatory authorities have been quick to recognize the need for specific support and have either published or initiated the creation of relevant guidance. Common misconceptions prevail however regarding sample size estimation and the choice of the most appropriate patient population to analyse while other areas such as equivalence margin specification and covariate adjustment have been neglected. This paper challenges some of the regulatory advice and the interpretation that others have made of this guidance with the aim of stimulating further debate. Copyright © 2003 John Wiley & Sons, Ltd.