Background: Pembro as 2L therapy was well tolerated and had clinically meaningful antitumor activity in Chinese pts with advanced melanoma at first analysis of the phase 1b KEYNOTE-151 study. At median FU of 7.9 mo, ORR was 16.7% (1 complete response [CR]; 16 partial response [PR]), DCR was 38.2%, median DOR was 8.4 mo, median PFS was 2.8 mo, and median OS was 12.1 mo. Any grade and grade 3/4 treatment-related adverse events (TRAEs) occurred in 84.5% and 8.7% of pts, respectively. However, long-term efficacy and safety of 2L pembro in Chinese pts with advanced melanoma are unknown. Results from 3 y of FU are presented. Methods: Eligible pts were ≥18 y, of Chinese descent, had histologically confirmed locally advanced or metastatic melanoma, and had disease progression on or after first line therapy. Pts received pembro 2 mg/kg IV Q3W for ≤35 cycles. Eligible pts who discontinued pembro with stable disease or better could receive ≤17 additional cycles of pembro (second course) upon disease progression. Primary endpoints were safety and ORR per RECIST 1.1 by blinded independent central review (BICR). Secondary endpoints included DOR and PFS per RECIST 1.1 by BICR; ORR, DOR, and PFS per irRECIST by BICR; and OS. ORR was based on the exact binomial method; DOR, PFS, and OS analyses used the Kaplan-Meier method. The full analysis set (FAS) comprised all allocated pts who received ≥1 pembro dose and had baseline data for relevant analyses. Results: Of 103 enrolled pts, 43% were male, 51% had PD-L1 positive tumors, and 80% had BRAF wild-type tumors; median age was 52 y (range, 22-77). The FAS had 102 pts. Median time from first dose to database cutoff was 44.6 mo (IQR, 39.1-46.2). At database cutoff (July 13, 2020), 14 pts (13.5%) had completed treatment, 89 pts (86.4%) had discontinued. Any grade TRAEs occurred in 88 pts (85.4%); most commonly hypothyroidism (n = 27; 26.2%), increased ALT (n = 24; 23.3%), and hypertriglyceridemia (n = 23; 22.3%), all grade 1/2. Grade 3-5 TRAEs occurred in 13 pts (12.6%). Immune-mediated AEs and infusion reactions occurred in 35 pts (34.0%); most commonly grade 1/2 hypothyroidism (n = 27; 26.2%). 3 pts discontinued pembro because of a TRAE; none died because of a TRAE. ORR per RECIST 1.1 was 17.6% (95% CI 10.8-26.4; 1 CR/17 PR); DCR was 38.2% (95% CI 28.8-48.4). Median DOR per RECIST 1.1 was 13.8 mo (range, 2.7-37.4+); 1 pt had response duration ≥36 mo. Median PFS per RECIST 1.1 was 2.8 mo (95% CI 2.7-3.5); 36-mo PFS rate was 5.0%. ORR, DOR, and PFS were similar per RECIST 1.1 and irRECIST. Median OS was 13.2 mo (95% CI 10.4-16.5); 36-mo OS rate, 22.3%.1 pt completed the first and second courses of pembro and had investigator-confirmed CR. Conclusions: 2L pembro was well tolerated and continued to provide durable responses and clinically meaningful antitumor activity in Chinese pts with advanced melanoma. These results further support use of 2L pembro in Chinese pts with advanced melanoma. Clinical trial information: NCT02821000