The primary variables are often systematically related to other influences apart from drug effect. For instance, there may be relationships to covariates such as health conditions or prognostic factors. When a $2{ imes}2$ crossover experiment for bioequivalence is designed, the statistical adjustment for the influence of covariates should be considered if some covariates influence the drug effect. Statistical inference for assessing average bioequivalence for a $2{ imes}2$ crossover design with covariates is given and an illustrated example is presented with discussion.