We investigate the long-run impact of college remedial education on students' academic performance and employment status. By implementing a Fuzzy Regression Discontinuity Design we show that attending remedial courses positively affects the probability to get a university degree, whereas no significant effect is found on labour market outcomes.