Immunotherapy targeting the programmed cell death-1 (PD-1) and programmed cell death-ligand 1 (PD-L1) pathway has emerged as an effective treatment for lung cancer patients. It is important to evaluate the practicality of PD-L1 testing in real-world practice. A total of 211 non–small cell lung cancer patients were enrolled to detect 5 driver mutations and PD-L1 status (22C3 and SP263 assays) and to evaluate the characteristics of PD-L1 expression and its predictive value of immunotherapy. The PD-L1 positive (≥1%) and strong positive (≥50%) rate by SP263 assay was 27.0% and 12.8%. The concordance rates between 2 PD-L1 assays while using 1%, 10%, 25%, and 50% positive tumor cells as the cutoffs were 76.8%, 81.5%, 90.5%, and 94.3%, respectively. Smokers and patients without known actionable driver mutation were more likely to present strong positive PD-L1 [adjusted hazard ratio, 5.00 (95% confidence interval—CI, 1.60-15.64); P=0.006 and 3.59 (95% CI, 1.25-10.33); P=0.018, respectively]. Higher levels of smoking were associated with higher PD-L1 expressions. None of the EGFR, ALK, HER2, or BRAFV600E-mutant nonsmokers displayed strong positive PD-L1 expression by SP263 assay. Among patients undergoing PD-1 checkpoint inhibitors therapy, high PD-L1 expression by SP263 was associated with a longer progression-free survival [adjusted hazard ratio, 0.15 (95% CI, 0.03-0.71); P=0.017]. In conclusion, our results suggest that PD-L1 status remains an important predictor of immunotherapy efficacy. The concordance between 22C3 and SP263 assays was greater at a higher cutoff level of positivity. Patients without known actionable driver mutation, along with smokers, particularly those having high smoking pack-years, were more likely to have strong PD-L1 expression.