In the last decade the European Employment Strategy strongly recommended reforms of active labour market policies, reforms that have generated a spread of evaluation exercises for most of European countries. This paper fills the gap in the literature concerning the Italian case, assessing the efficacy of Public Employment Services (PESs) -after the reforms of 1997, 2000, 2003- in increasing the unemployment to employment transition probabilities, through matching techniques. Exploiting the longitudinal dimension of the Labour Force Survey data we design an evaluation structure that allows observing outcomes in both the short (at most 3 months) and the long run (at most 12 and 15 months). In this framework, PES users show a lower probability of finding a job in the short term, because of a lock-in effect, while in the long term this probability turns out to be positive. We also show that PES effects in the long term are much less pronounced when considering as outcome variable the probability of finding a permanent job, a proxy for the quality of the job, suggesting that PES impacts are to a large extent driven by the use of temporary contracts. Furthermore, to deal with issues related to selection on unobservables we carry out two different sensitivity analysis, which confirm our baseline findings.