Aims. To compare short-term and long-term outcomes of preoperative endoscopic biliary drainage (EBD) and percutaneous biliary drainage (PBD) in patients with Klatskin tumor undergoing curative surgery. Methods. We conducted a search of electronic information sources to identify all studies comparing EBD and PBD in patients with Klatskin tumor undergoing curative surgery. We used the Newcastle-Ottawa Scale to assess the risk of bias observational studies. Random-effects or fixed-effects modeling was applied as appropriate to calculate pooled outcome data. Results. We identified 9 observational studies, enrolling a total of 1436 patients. The patients in the PBD group had more advanced disease than those in EBD group in terms of Bismuth-Corlette classification and tumor classification. EBD was associated with higher risks of postprocedural complications (odds ratio [OR] =2.24, P = .001), conversion to another drainage method (OR =11.16, P < .00001), cholangitis (OR = 4.58, P < .0001), and pancreatitis (OR = 8.90, P = .009) than PBD; there was no difference between the 2 methods in terms of technical success (OR = 0.79, P = .50) and tube dislocation (OR = 0.81, P = .54). Regarding the postoperative outcomes, there was no difference in terms of 30-day mortality (OR = 0.61, P = .16) and major postoperative complications (OR = 0.60, P = .06). Regarding the long-term outcomes, EBD was associated with lower risks of seeding metastasis (OR = 0.46, P = .0004) and 5-year recurrence (OR = 0.72, P = .010), and better 5-year survival (OR = 1.62, P = .001). Conclusions. EBD may be associated with higher procedure-related complications compared with PBD as a preoperative biliary drainage method in patients with Klatskin tumor undergoing curative surgery. The available evidence on long-term oncological and survival outcomes are subject to confounding by indication, and high-quality randomized controlled trials are required for definite conclusions.