Previous research on Italian higher education showed that social origin affected students' academic progression and results in the 20th century. In this paper we examine the role of student employment - i.e. working during university - in the reproduction of social inequality in academic outcomes. In the first part, we review previous research results in the US, UK and Italy and discuss several competing hypotheses. In the second part, we use data from the Italian Longitudinal Household Survey (ILFI) to study a) the relation between student employment and academic outcomes; b) the relation between social origin and student employment, and c) the mediating effect of student employment in the relation between social origin and academic outcomes. Bivariate analysis and multinomial logistic regression models show that full-time students are more likely to graduate on time than working-students, but only high-intensity work has a detrimental effect on dropping out. Social origin affects the probability of being a high-intensity worker, but not the likelihood of being a low-intensity worker. Finally, results from a non-linear decomposition analysis suggest that the overall role of student employment in the reproduction of inequality in higher education is low, while the most important variable is the type of highschool attended (especially lyceum vs non-academic).