This paper exploits detailed offender-level data from the public prosecutor service to analyze whether a reform aimed at reducing recidivism was effective. The objective of this reform was to enhance deterrence by improving the certainty and celerity of punishment. More specifically, it (1) stimulated prosecutors to intensify the use of alternative dispositions (such as imposing settlements) in criminal cases that would otherwise have been dismissed, (2) reduced processing times by improving the collaboration between police and prosecutors, and (3) encouraged tailored decisions. Using a difference-in-differences approach, I exploit variation in when and where the district attorney reform was introduced to quantify its effect on recidivism rates. In doing so, I am able to compare otherwise similar individuals, who committed similar crimes, but who underwent different procedures. I further use an event study to assess the evolution of relative recidivism rates, and explore possible mechanisms including immediacy and certainty of punishment. The data reveal that after the reform, the prospects of recidivism linked to local crimes targeted by the reform decreased by 5 percentage points, a 26% reduction over the sample mean.