Abstract
The two one-sided tests procedure of Schuirmann (1987) has become the standard technique for assessing bioequivalence of two formulations of the same drug. The power of this procedure is examined to determine the sample size required to control the power. A simple approximation to the required sample size is presented.