Résumé : Background and Aims: Preoperative biliary drainage with self-expanding metal stents (SEMSs) brings liver function within an acceptable range in preparation for neoadjuvant therapy (NATx) and provides relief of obstructive symptoms in patients with pancreatic cancer. We compared fully-covered SEMSs (FCSEMSs) and uncovered SEMSs (UCSEMSs) for sustained biliary drainage before and during NATx. Methods: Patients with pancreatic cancer and planned NATx who need treatment of jaundice and/or cholestasis before pancreaticoduodenectomy were randomized to FCSEMSs versus UCSEMSs. The primary endpoint was sustained biliary drainage, defined as the absence of reinterventions for biliary obstructive symptoms, and was assessed from SEMS placement until curative intent surgery or at 1 year. Results: The intention-to-treat population included 119 patients (59 FCSEMSs, 60 UCSEMSs). Sustained biliary drainage was equally successful with FCSEMSs and UCSEMSs (72.2% vs 72.9%, noninferiority P =.01). Reasons for FCSEMS and UCSEMS failure differed significantly between the groups and included tumor ingrowth in 0% versus 16.7% (P <.01), and stent migration in 6.8% versus 0% (P =.03), respectively. Serious adverse event rates related to stent placement were not significantly different in both groups (23.7% [14/59] vs 20.0% [12/60], P =.66), as were acute cholecystitis rates when the gallbladder was in situ (9.3% [4/43] vs 4.8% [2/42], P =.68) for FCSEMSs and UCSEMSs, respectively. In our study, independent of stent type, predictors of reinterventions were 4-cm stent length and presence of the gallbladder. Conclusion: FCSEMSs and UCSEMSs provide similar preoperative management of biliary obstruction in patients with pancreatic cancer receiving NATx, but mechanisms of stent dysfunction depend on stent type, stent length, and presence of the gallbladder. (Clinical trial registration number: NCT02238847.)