Background: Pembrolizumab (PD-1 inhibitor) and cetuximab (EGFR inhibitor) are active as single agents and in combination with cytotoxic chemotherapy for recurrent or metastatic head and neck squamous cell carcinoma (HNSCC). Given each drug's single agent activity and unique mechanism of action, we aimed to evaluate the anti-tumour activity of PD-1 blockade with EGFR inhibition in recurrent or metastatic HNSCC. Methods: This study is an open-label, non-randomised, multi-arm, phase 2 trial done at four academic centres in the USA. Participants were required to have platinum-resistant or platinum-ineligible, recurrent or metastatic HNSCC, be at least 18 years old, have an Eastern Cooperative Oncology Group performance status 0–1, have measurable disease per Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1, and to have received no previous immunotherapy or EGFR inhibition. All participants received pembrolizumab 200 mg intravenously every 3 weeks, combined with an initial loading dose of cetuximab 400 mg/m2 intravenously followed by 250 mg/m2 intravenously weekly (21 day cycle). The primary endpoint was overall response rate defined as the proportion of participants with a partial or complete responses (per RECIST version 1.1) by 6 months in the intention-to-treat population. The safety population included all participants who received at least one dose of pembrolizumab. Herein, the final analysis of cohort 1 (no previous PD-1, PD-L1, or EGFR inhibition for recurrent or metastatic HNSCC) is reported. Three additional cohorts (two for participants with a previous response to immunotherapy followed by relapse or progression, with or without previous cetuximab exposure, and one for cutaneous HNSCC) will be reported separately once fully accrued. This study is registered with ClinicalTrials.gov, NCT03082534, and remains open as the three additional cohorts are actively accruing participants. Findings: Between March 22, 2017, and July 16, 2019, 33 participants were enrolled to cohort 1. All 33 participants received at least one dose of pembrolizumab. Median follow-up duration was 7·3 months (IQR 3·9–10·9). By 6 months, the overall response rate was 45% (95% CI 28–62), with 15 of 33 participants achieving a partial response. The most common grade 3–4 treatment-related adverse event was oral mucositis (three [9%] of 33 participants), and serious treatment-related adverse events occurred in five (15%) participants. No treatment-related deaths occurred. Interpretation: Pembrolizumab combined with cetuximab shows promising clinical activity for recurrent or metastatic HNSCC, and merits further investigation. Funding: Merck Sharp & Dohme.