This paper investigates the empirical relevance of a new framework for monetary policyanalysis in which decision makers are allowed to weight differently positive and negative deviations of inflation and output from the target values. The specification of the central bank objective is general enough to nest the symmetric quadratic form as a special case, thereby making the derived policy rule potentially nonlinear. This forms the basis of our identification strategy which is used to develop a formal hypothesis testing for the presence of asymmetric preferences. Reduced-form estimates of postwar US policy rules indicate that the preferences of the Fed have been highly asymmetric in both inflation and output gaps, with the asymmetries on the latter becoming relatively more pronounced during the post-79 tenures.