Statistical interpretations in a bioequivalence trial are considered and studied when the missing observations occurred in 2 × 2 crossover experiment. Patel (1985) suggested the approximate test procedures for carryover effect and drug effect in 2 × 2 crossover design when some of data are missing in the second period. A modified Patel method is newly pro-posed to the bioequivalence trial and it is compared with the current method through the simulation study.