To better understand unemployment dynamics it is key to assess the role played by job creation and job destruction. Although the U.S. case has been studied extensively, the importance of job finding and employment exit rates to unemployment variability remains unsettled. The aim of this paper is to contribute to this debate by adopting a novel lens, wavelet analysis. We resort to wavelet analysis to unveil time- and frequency-varying features regarding the contribution of the job finding and job separation rates for the U.S. unemployment rate dynamics. Drawing on this approach, we are able to reconcile some apparently contradictory findings reported in previous literature. We find that the job finding rate is more influential for the overall unemployment behavior but the job separation rate also plays a critical role, especially during recessions.