This paper evaluates the forecast performance of boosting in comparison to the forecast combination schemes and dynamic factor models presented in Stock and Watson (2006). We find that boosting is a serious competitor for forecasting US industrial production.