Previously, the connection between bioequivalence tests and their asso
ciated confidence intervals was not well understood. In fact, the alph
a-level two one-sided tests approach was thought to be associated with
a 1 - 2 alpha confidence interval. In this paper, we build up connect
ions which allow us to construct 1 - alpha confidence intervals from a
lpha-level tests and vice versa. When applied to the alpha-level two o
ne-sided tests, the resultant 1 - alpha confidence intervals are prope
rly contained in Westlake's 1 - alpha symmetric interval (1976). Our a
pproach is readily generalized to different settings, including the no
nparametric setting, the ratio parameter setting, and the repeated con
fidence interval setting.