We explore the use of univariate low-frequency filters in macroeconomic forecasting. This amounts to targeting only specific fluctuations of the time series of interest. We show through simulations that such approach is warranted and, using US data, we confirm empirically that consistent gains in forecast accuracy can be obtained in comparison with a variety of other methods. There is an inherent arbitrariness in the choice of the cut-off defining low and high frequencies, but we show that some patterns characterize the implied optimal (for forecasting) degree of smoothing of the key macroeconomic indicators we analyze. For most variables the optimal choice amounts to disregarding fluctuations well below the standard business cycle cut-off of 32 quarters while generally increasing with the forecast horizon; for inflation and variables related to housing this cut-off lies around 32 quarters for all horizons, which is below the optimal level for federal spending.