Background: Cytoreductive surgery (CS) including cytoreductive nephrectomy and metastasectomy provides clinical benefits to patients with MRCC. However, CS has not been evaluated in the setting of immune checkpoint (IC) therapy. We performed a pre-surgical/biopsy trial to evaluate biological and clinical activity of nivo +/- (bev or ipi) in patients with MRCC. Methods: In this open-label, randomized trial (NCT02210117), patients with MRCC w/o prior IC therapy and anti-VEGF therapy were randomized 2:3:2 to receive nivo (3mg/kg q2wks x3), nivo + bev (10mg/kg q2wks x3) or nivo + ipi (1mg/kg q3wks x2), followed by CS or Bx, and then nivo maintenance therapy up to 2 yrs. Clinical response per RECIST criteria was assessed at ≥12 wks. Pre- and post-treatment blood and tumors were obtained for correlative studies. Results: All of 105 patients have been accrued and 104 are evaluable for response. For all patients, best overall response (BOR = complete response [CR] + partial response [PR]) including surgery effect was 55% nivo, 44% nivo + bev, 43% nivo + ipi. Median PFS (95% confidence interval) was 14.5 months (5.5, not reached [NR]) nivo, 7.6 (4.8, 8.9) months nivo + bev, 7.5 (2.0, 12.4) months nivo + ipi. Overall survival (OS) at one year was 86% nivo, 73% nivo + bev, 83% nivo + ipi. Grade 3 or higher toxicities related to therapy were 38% for nivo, 42% for nivo + bev (including 18% hypertension), and 47% for nivo + ipi. For patients with CS, BOR including surgery effect was: 86% nivo, 88% nivo + bev, and 69% nivo + ipi. Median PFS was 17.3 months (6.1, NR) nivo, 7.6 (5.7, 10.3) months nivo + bev, 8.9 (2.9, 23.0) months nivo + ipi. OS at one year was 100% nivo, 94% nivo + bev, 92% nivo + ipi. Median OS has not yet reached with a median follow-up of 24.6 months. Immune and gene profiling analyses demonstrate: 1) tumor infiltrating CD8 T cells correlate with clinical responses to nivo or nivo + bev, but not to nivo + ipi; 2) tumor IFN pathway gene expression correlates with responses; and 3) PD-L1 status, tumor mutation or mutation burden, neoantigens did not correlate with response. Conclusions: IC therapy plus CS is safe and beneficial to patients with MRCC and therefore, warrants testing, along with a few correlative biomarkers, in a larger phase 3 trial. Clinical trial information: NCT 02210117.