Monalizumab (M), targeting the natural killer group 2A (NKG2A) receptor, has limited activity as monotherapy in recurrent/metastatic (R/M) squamous cell carcinoma of the head and neck (SCCHN). Preliminary data of M and durvalumab (D) have shown encouraging activity in other tumor types. The UPSTREAM trial was an umbrella trial of targeted therapies and immunotherapy for R/M SCCHN. The immunotherapy 2 (I2) cohort was a phase II, randomized, open-label substudy evaluating the efficacy of D + M versus physician's choice (control). Patients non-eligible for the biomarker-driven cohorts and pretreated with PD(L)1, were included in the I2 cohort. The primary endpoint was the objective response rate (RECIST version 1.1) during the first 16 weeks. Sixty-six patients with R/M SCCHN were included in the I2 cohort, of whom 60 were assessable (D + M: n = 42, control: n = 18): median age 62 years; 87% with two or three previous lines of treatment. In the D + M arm, one partial response (PR) was recorded, and stable disease (SD) was observed in 11 (26%). One PR was reported in the control arm and SD in 8 (44%). The median progression-free survival (PFS) was 2.0 and 3.1 months in the D + M arm and control arm, respectively. The median overall survival (OS) was 4.3 months (95% confidence interval 3.3-8.9 months) and 8.0 months (95% confidence interval 3.1-14.9 months) in the D + M and control arms, respectively. In the D + M arm, 4 (9%) patients reported grade ≥3 treatment-related adverse events. The I2 substudy failed to demonstrate an activity of D + M in heavily pretreated patients with SCCHN previously exposed to anti-PD(L)1. No benefit was seen in PFS and OS.