Background: This phase 1 study evaluated the safety and efficacy of the immune checkpoint inhibitor nivolumab (monotherapy or combined with chemotherapy) in Japanese patients with biliary tract cancer (BTC). Methods: In the monotherapy cohort (N = 30), patients with unresectable/recurrent BTC refractory or intolerant to gemcitabine-based treatment regimens received nivolumab monotherapy (240 mg, 2-week intervals). In the combined therapy cohort (N = 30), chemonaïve patients with unresectable/recurrent BTC received nivolumab (240 mg, 2-week intervals) plus cisplatin-gemcitabine chemotherapy. The primary objective was tolerability and safety. Secondary efficacy endpoints included overall survival (OS), progression-free survival (PFS) and objective response rate (ORR) (central assessment). Results: The most frequently reported nivolumab-related adverse events were decreased appetite (5/30, 17%), malaise (4/30, 13%), and pruritus (4/30, 13%) in the monotherapy cohort, and malaise (8/30, 27%) and decreased appetite (7/30, 23%) in the combined therapy cohort. Median duration of follow-up for efficacy (Table) was 5.1 months (mo; monotherapy cohort) and 8.2 mo (combined therapy cohort). In the monotherapy cohort, median OS and median PFS were longer in patients with programmed death-ligand 1 (PD-L1) ≥ 1% (including the patient who responded) than in those with PD-L1 < 1% (Table). In the combined therapy cohort, the ORR was higher, but median OS and median PFS were lower, in patients with PD-L1 ≥ 1% than in those with PD-L1 < 1% (Table). Conclusions: Nivolumab was well tolerated, with a manageable safety profile and signs of clinical activity, in this phase 1 study in patients with unresectable/recurrent BTC. PD-L1 expression status may predict clinical activity of nivolumab in advanced BTC. Clinical trial information: JapicCTI-153098.