We analyse how the labour market implicitly evaluates Italy’s higher education system by estimating differences in employment and earnings across universities. We use our estimates to produce three rankings of universities based, respectively, on employment, earnings and employment-weighted earnings. By controlling for a large set of covariates, we isolate each university effect on employment and earnings from additional components influencing graduates' labour market outcomes, namely the university’s field of specialization, the graduates' observable characteristics and their local labour markets. To account for the latter, we include graduates' employment rate in the region of residence among the covariates but we instrument it with prior residence in order to correct for endogenous sorting. We discuss pros and cons of our methodology and compare our results with other available university rankings.