
This study aims to understand how custody hearings influence pretrial detention. These hearings were implemented in Brazil in 2015 as an experimental project to improve judicial decision-making. From the outset, their primary goal was to enhance the use of precautionary measures other than imprisonment, thereby reducing the number of pretrial detainees and limiting detention to strictly necessary cases. To assess whether custody hearings have effectively led to more release decisions than pretrial detention orders, we analyzed 4.374 judicial decisions from drug trafficking cases closed in 2019. Among them, 2.505 involved custody hearings, while 1.869 decisions were made outside of hearings (in chambers). Three binary logistic regression models were constructed, using the decision for detention or release as the dependent variable. Contrary to initial expectations, custody hearings do not increase the likelihood of release decisions, particularly when the defense is conducted by a private attorney. In both custody hearings and in-chamber decisions, the main factor determining detention is the indication of involvement with criminal organizations. Finally, the detainee’s profile influences outcomes differently in each context. In chamber decisions, Black men are more likely to be detained. In custody hearings, unemployed men face a higher likelihood of detention, which shows how the profile of the detainee can influence the custody hearings.