The paper introduces a new measure of educational mismatch at the firm level, constructed by merging firm and individual data at the sector-firm size-year level. This measure captures both the intensity of mismatch and its type – whether overeducation, undereducation, or a mix of the two. We assess the role of human resource practices in reducing the intensity of educational mismatch in Italian firms by estimating econometric models that control for a rich set of firm characteristics, as well as year and industry-region fixed effects. Firm-fixed effects and instrumental variable models complement the analysis. Findings indicate that the use of private recruitment agencies, on-the-job training, and structured supervision is associated with a reduction in mismatch intensity. The impact of other practices varies by mismatch type: higher job turnover rates correlate with lower undereducation but increased overeducation, while second-level bargaining increases undereducation and reduces overeducation.