Fixed doses at 200 mg of pembrolizumab or 2 mg/kg every 3 weeks are the standard dosages for first- and second-line treatment of non-small-cell lung cancer (NSCLC); however, in clinical practice, patients with NSCLC may receive lower doses of pembrolizumab due to drug product availability or economic factors. To date, the comparative effectiveness and safety of the standard dose and lower doses of pembrolizumab in these patients still remains limited. We conducted a retrospective cohort study by analyzing electronic medical records data from the largest multi-institutional hospital system in Taiwan. Advanced NSCLC patients newly receiving pembrolizumab with or without chemotherapy were included. Patients were classified into: (1) the standard-dose group (≥2 mg/kg), and (2) the low-dose group (<2 mg/kg). We applied inverse probability of treatment weighting (IPTW) to compare the overall survival (OS) and immune-related adverse events (irAEs) between the two treatment groups, and to evaluate the minimum clinically effective dose of pembrolizumab. We included a total of 147 NSCLC patients receiving standard-dose pembrolizumab (mean [range] age: 63.7 [58.0−73.0] years; male: 62.6%; mean [range] body weight: 60.5 [58.0−73.0] kg) and 95 patients receiving low-dose pembrolizumab (mean [range] age: 62.0 [50.0−68.8] years; male: 64.2%; mean [range] body weight: 63.9 [55.0−73.8] kg). After IPTW adjustments, the median OS was similar for both the standard-dose and low-dose pembrolizumab groups (19.3 vs. 14.3 months, log-rank p = 0.15). Also, the rate for all classes of irAEs was similar for both groups. We found that patients with a pembrolizumab dose ≥1.8 mg/kg were associated with better OS than those receiving <1.8 mg/kg. Our findings suggested no significant difference in OS and irAEs between patients receiving pembrolizumab ≥2 mg/kg and <2 mg/kg in clinical practice. A pembrolizumab dose ≥1.8 mg/kg may be the clinically most efficient dose.