The recently published EMA Bioequivalence Guidelines (2010) stipulates the application of fixed effect ANOVA models should be used when proving bioequivalence. In addition, the CHMP PK Working Party (2011) states that data not relevant to the treatment comparison of interest should be removed. This contrasts with the commonly applied approach where all available data is used within a mixed model. The issue is not only for BE studies, but potentially influential on DDI or TQT studies which are crossover designs with 3 or more treatments. This roundtable discussion proposes to compare the approaches presented in the EMA and FDA guidance (2003) documents when proving bioequivalence and the application of both within a global setting.