This article presents the results of an empirical exercise aimed at estimating employment effects of public vocational training courses completed by unemployed in an Italian region (Piedmont) in 2015. Using propensity score matching techniques we find a positive causal effect of the training: one year after the end of the courses the employment rate is about 14 percentage points higher in the treated group compared to the control group. Estimated effects are heterogeneous being more effective the training courses enhancing specialized competencies. Moreover, the impact of the program is higher for younger, short-term unemployed, and those attending longer training courses. The effects do not vanish in the medium term and the results have been proved to be robust to a list of sensitivity checks.