There are two main hypotheses as to dynamic trends of unemployment in the literature. The first of these is the natural rate hypothesis put forth by Friedman (1968) and Phelps (1968), and the second proposition on unemployment is known as the "hysteresis" hypothesis. According to this latter hypothesis developed through the support of Blanchard and Summers (1986, 1987), Layard et al, (1991) and Barron (1988), the shocks have permanent effects on unemployment levels due to rigidities in the labor market. In other words, the unemployment rate tends to increase and does not get back to its previous level after a shock in the economy. This study examines the validity of unemployment hysteresis in the case of EU and Turkey by using quarterly data over the period 2001.Q12015.Q4, i) for each country with time-series, and ii) for selected EU countries with panel data analysis. Since the analysis incorporates both frequency and time dimensions, the existence of unemployment hysteresis is estimated by techniques based on Fourier-ADF and Fourier IPS approaches. According to the findings of the study, unemployment rates follow a linear pattern except for Italy, Spain and Estonia. The panel data analysis of the unemployment rates in the EU demonstrates a linear structure and unemployment hysteresis is proven to be valid across EU countries.