This article analyzes the hysteresis hypothesis in the unemployment rates of the four “French overseas regions” (Guadeloupe, Martinique, Guyana, Reunion) [FORs] over the period 1993-2008. We use standard univariate and panel unit root tests, among them Choi(2006) and Lopez (2009) that account for cross-sectional dependence and have improved performance when the number of countries and the time dimension of the data are limited. Our results cannot reject the null hypothesis of a unit root and so find evidence supporting hysteresis in the unemployment rates for the FORs.