© 2015 Elsevier Ireland Ltd. Objectives: The immune effects of EGFR tyrosine kinase inhibitors (EGFR-TKIs) are poorly understood. Identifying immune biomarkers could guide patient selection and optimisation of EGFR-TKI-immunotherapy combinations. Materials and methods: 33 patients with NSCLC treated with an EGFR-TKI were prospectively enrolled. Peripheral blood mononuclear cells were collected pre-treatment, and after 1, 3 and 8 weeks. Flow cytometry was used to identify immune cell subsets, including PD-1 and PD-L1 expressing T cells. Immune parameters were correlated with clinical outcomes. Results: Compared to healthy donors (n = 10), patients had higher pre-treatment proportions of proliferating and PD-L1+CD3+ T cells (p <0.001). Compared to patients with an EGFR mutation (n = 12), patients without a known mutation (n = 21) had higher proportions of proliferating CD4+ and PD-L1+CD3+ T cells (p = 0.03).There was a significant increase in PD-L1+ T cells after 1 week of EGFR-TKI in patients whose disease progressed compared to non-progressors. Patients with higher PD-L1+CD3+ T cells at 1-week were more likely to progress (OR 30.3, p <0.01) and had shorter PFS (1.6 vs. 8.8 m; p <0.01) and OS (3.8 vs 23.2 m; p <0.001) than those with fewer PD-L1+CD3+ T cells. On multivariate analysis, high PD-L1+CD3+ T cells was the only independent predictor for PFS (HR 3.7, p = 0.01), while for OS independent predictors were high PD-L1+CD3+ T cells (HR 6.5, p <0.01) and EGFR-negative status (HR 3.3, p = 0.04). Conclusions: There was a significant correlation between PD-L1 expression on peripheral T cells and clinical outcomes in EGFR-TKI-treated NSCLC. This warrants further validation as a blood-based biomarker that may identify candidates for PD-1 inhibitors or immunotherapy-EGFR-TKI combinations.