High interindividual variability (IIV) of the clinical response to epidermal growth factor receptor (EGFR) inhibitors such as osimertinib in non-small-cell lung cancer (NSCLC) might be related to the IIV in plasma exposure. The aim of this study was to evaluate the exposure–response relationship for toxicity and efficacy of osimertinib in unselected patients with advanced EGFR-mutant NSCLC. This retrospective analysis included 87 patients treated with osimertinib. Exposure–toxicity analysis was performed in the entire cohort and survival analysis only in second-line patients (n = 45). No significant relationship between occurrence of dose-limiting toxicity and plasma exposure was observed in the entire cohort (p = 0.23, n = 86). The median overall survival (OS) was approximately two-fold shorter in the 4th quartile (Q4) of osimertinib trough plasma concentration (>235 ng/mL) than in the Q1–Q3 group (12.2 months [CI95% = 8.0–not reached (NR)] vs. 22.7 months [CI95% = 17.1–34.1]), but the difference was not statistically significant (p = 0.15). To refine this result, the exposure–survival relationship was explored in a cohort of 41 NSCLC patients treated with erlotinib. The Q4 erlotinib exposure group (>1728 ng/mL) exhibited a six-fold shorter median OS than the Q1–Q3 group (4.8 months [CI95% = 3.3-NR] vs. 22.8 months (CI95% = 10.6–37.4), p = 0.00011). These results suggest that high exposure to EGFR inhibitors might be related to worse survival in NSCLC patients.