PURPOSE: To study the impact of transarterial Yttrium-90 radioembolization (TARE) in combination with second-line systemic chemotherapy for colorectal liver metastases (CLM). METHODS: In this international, multicenter, open-label phase III trial, patients with CLM who progressed on oxaliplatin- or irinotecan-based first-line therapy were randomly assigned 1:1 to receive second-line chemotherapy with or without TARE. The two primary end points were progression-free survival (PFS) and hepatic PFS (hPFS), assessed by blinded independent central review. Random assignment was performed using a web- or voice-based system stratified by unilobar or bilobar disease, oxaliplatin- or irinotecan-based first-line chemotherapy, and KRAS mutation status. RESULTS: Four hundred twenty-eight patients from 95 centers in North America, Europe, and Asia were randomly assigned to chemotherapy with or without TARE; this represents the intention-to-treat population and included 215 patients in the TARE plus chemotherapy group and 213 patients in the chemotherapy alone group. The hazard ratio (HR) for PFS was 0.69 (95% CI, 0.54 to 0.88; 1-sided P = .0013), with a median PFS of 8.0 (95% CI, 7.2 to 9.2) and 7.2 (95% CI, 5.7 to 7.6) months, respectively. The HR for hPFS was 0.59 (95% CI, 0.46 to 0.77; 1-sided P < .0001), with a median hPFS of 9.1 (95% CI, 7.8 to 9.7) and 7.2 (95% CI, 5.7 to 7.6) months, respectively. Objective response rates were 34.0% (95% CI, 28.0 to 40.5) and 21.1% (95% CI, 16.2 to 27.1; 1-sided P = .0019) for the TARE and chemotherapy groups, respectively. Median overall survival was 14.0 (95% CI, 11.8 to 15.5) and 14.4 months (95% CI, 12.8 to 16.4; 1-sided P = .7229) with a HR of 1.07 (95% CI, 0.86 to 1.32) for TARE and chemotherapy groups, respectively. Grade 3 adverse events were reported more frequently with TARE (68.4% v 49.3%). Both groups received full chemotherapy dose intensity. CONCLUSION: The addition of TARE to systemic therapy for second-line CLM led to longer PFS and hPFS. Further subset analyses are needed to better define the ideal patient population that would benefit from TARE.