In this paper, we empirically assess the Taylor principle in the long run with a small‐scale, trivariate structural vector autoregression. Using U.S. data spanning the period from 1959Q1 to 2010Q2 and different measures of inflation and the output gap, we find that the Taylor principle is supported in the long run for the post‐1979 era, but the principle is unlikely to hold in the long run for the pre‐1979 era. (JEL E52, E58)