In their classic paper, Box and Cox (1964) demonstrated how a dependent variable could be transformed to satisfy simultaneously, assumptions implicit in the analysis of linear models. For the class of analyses in which the response of interest is positive and where no transformation of independent variables is considered, a SAS macro is presented which makes the estimation of this power transformation relatively easy. The use of the macro is illustrated with data taken from the original article of Box and Cox. Its application in investigating the effects of such a transformation on non-additivity is also demonstrated.