Background: There is no standard treatment beyond progression on first-line pembrolizumab (P) in patients with R/M HNSCC. Vascular endothelial growth factor (VEGF) modulates anti-tumor immunity via proliferation of inhibitory T regulatory and myeloid-derived suppressor cells, inhibition of dendritic cell maturation and suppression of effector T cell responses. Efficacy data from other disease sites and from early-phase trials done in HNSCC suggests promising clinical activity of VEGF and PD-1 inhibitor combination therapy. We have thus designed a Phase II/III, randomized, multi-arm trial to evaluate the efficacy of atezolizumab (A) and bevacizumab (B) beyond progression on P in patients with R/M HNSCC. Methods: EA3202 is enrolling patients with CPS ≥1% R/M HNSCC previously treated with P either alone or in combination with other checkpoint inhibitors without progression for at least 12 weeks. In phase II, patients will be randomized to one of three treatment arms: platinum doublet chemotherapy (C) plus cetuximab (E) for six cycles followed by E maintenance (control arm), C + B for six cycles followed by B maintenance, or A + B. Therapy in each arm will continue until progression, toxicity, or for a total period of two years. Patients will be stratified by p16 status, CPS score (≥ 20 vs. < 20), distant metastases (M0 vs M1), and disease progression pattern with first-line P (while on P vs. after discontinuation). 216 patients will be enrolled. The primary endpoint for phase II is progression-free survival (PFS). Each experimental arm will be compared to the control arm at a one-sided alpha level of 0.10. Pre-specified rules will be used to pick a winner among the two experimental arms, which will then advance to phase III against the control arm. 214 patients will be enrolled in phase III for a total sample size of 430 patients. The primary endpoint for phase III is overall survival (OS) which will be compared using a stratified log-rank test. Secondary endpoints include OS in the two phase III arms among the CPS ≥ 20 cohort and comparison of grade 3 or higher treatment-related adverse events between the two arms. A total of 277 OS events will be required for full information for phase III evaluation to give a nominal ̃90% power at one-sided alpha level 0.0125. 144 patients enrolled in phase II (from the control and selected experimental arms) will be included in the phase III analysis. With this design the overall study power is at least 80% under a true 40% hazard reduction for PFS and 35% hazard reduction for OS. Blood and tumor tissue will be banked. EA3202 was activated in December 2021. It is the first randomized study to compare systemic treatments for R/M HNSCC in the immunotherapy era and will define the best second-line treatment approach beyond P. Clinical trial information: NCT05063552.