The usefulness of the derived neutrophil-to-lymphocyte ratio (dNLR) and its dynamics before/after durvalumab consolidation therapy to predict safety or efficacy remains unclear. We retrospectively reviewed patients with locally advanced non-small cell lung cancer treated with durvalumab consolidation therapy after chemoradiotherapy (D group) or chemoradiotherapy alone (non-D group) at multiple institutions. We investigated the association between dNLR, or its dynamics, and pneumonitis, checkpoint inhibitor-related pneumonitis (CIP), irAEs, and efficacy. Ninety-eight and fifty-six patients were enrolled in the D and non-D groups, respectively. The dNLR at baseline was significantly lower in patients who experienced irAEs or CIP than in those who did not. The low dNLR group, 28 days following durvalumab consolidation therapy (dNLR28 ≤ 3), demonstrated longer progression-free survival (PFS) and overall survival (OS) than the high dNLR group (dNLR28 > 3) (PFS, hazard ratio [HR] 0.44, 95% confidence interval [CI] 0.22–0.88, p = 0.020; OS, HR 0.39, 95% CI 0.16–0.94, p = 0.037). Among patients with high dNLR at baseline (dNLR > 3), the dNLR28 ≤ 3 group showed longer PFS than the dNLR28 > 3 group (p = 0.010). The dNLR is a predictive factor for irAEs and CIP in patients receiving durvalumab consolidation therapy. The dNLR at 28 days after durvalumab consolidation therapy and its dynamics predict favorable outcomes.