This paper proposes a new econometric estimation method for analyzing the probability of leaving unemployment using uncompleted spells from repeated cross-section data, which can be especially useful when panel data are not available. The proposed method-ofmoments- based estimator has two important features: (1) it estimates the exit probability at the individual level and (2) it does not rely on the stationarity assumption of the inflow composition. We illustrate and gauge the performance of the proposed estimator using Spanish Labor Force Survey data, and analyze the changes in distribution of unemployment between the 1980s and 1990s during a period of labor market reform. We find that the relative probability of leaving unemployment for the short-term unemployed compared to the long-term unemployed becomes significantly higher in the 1990s.