A number of methods have been proposed for statistical analysis of bioequivalence trials. A new method is proposed and compared to three previously suggested techniques. This comparison clarifies the relationship among the four methods and yields practical suggestions regarding statistical tests.