This paper investigates post-war U.S. business cycles using quarterly data. The study focuses on the rapid fluctuations in real output and other macroeconomic variables, distinguishing them from slow-moving components. Key findings include the positive association between output fluctuations and variations in hours worked, with similar percentage magnitudes. High-frequency variations in capital and productivity are weakly associated with output fluctuations, contrasting with slow-moving components. Cyclical investment varies more in percentage terms than cyclical output, while cyclical consumption varies less. The study documents systematic deviations from neoclassical growth theory. The authors use a decomposition method to separate growth and cyclical components, with the growth component assumed to vary smoothly over time. The smoothing parameter is chosen based on prior knowledge and statistical considerations. The results show that the growth component varies between 2.3 and 4.9 percent annually. The study emphasizes the importance of using the same smoothing parameter for all series and highlights the stability of certain economic variables. The paper also discusses the serial correlation properties of the data series, noting differences between real output and other variables. The results indicate considerable persistence in the rapidly varying component of output. The study concludes that the decomposition method provides a useful framework for analyzing business cycles and highlights the importance of considering both variability and covariability in economic data.This paper investigates post-war U.S. business cycles using quarterly data. The study focuses on the rapid fluctuations in real output and other macroeconomic variables, distinguishing them from slow-moving components. Key findings include the positive association between output fluctuations and variations in hours worked, with similar percentage magnitudes. High-frequency variations in capital and productivity are weakly associated with output fluctuations, contrasting with slow-moving components. Cyclical investment varies more in percentage terms than cyclical output, while cyclical consumption varies less. The study documents systematic deviations from neoclassical growth theory. The authors use a decomposition method to separate growth and cyclical components, with the growth component assumed to vary smoothly over time. The smoothing parameter is chosen based on prior knowledge and statistical considerations. The results show that the growth component varies between 2.3 and 4.9 percent annually. The study emphasizes the importance of using the same smoothing parameter for all series and highlights the stability of certain economic variables. The paper also discusses the serial correlation properties of the data series, noting differences between real output and other variables. The results indicate considerable persistence in the rapidly varying component of output. The study concludes that the decomposition method provides a useful framework for analyzing business cycles and highlights the importance of considering both variability and covariability in economic data.