Background: Since the publication of KEYNOTE-189 study, carboplatin/pemetrexed/pembrolizumab (carbo/pem/pembro) has become a common frontline regimen for advanced non-small cell lung cancer (NSCLC). In KEYNOTE-189, carboplatin was administered for a maximum of 4 cycles. To date, the number of carboplatin cycles administered in the real-world setting remains unclear. We analyzed factors predicting the receipt of carboplatin beyond 4 cycles (CB4). Methods: We interrogated a population-based, real-world database by Flatiron for NSCLC patients with no actionable EGFR/ALK who received carbo/pem/pembro from March to July 2019 (pre-pandemic period) and from March to July 2020 (pandemic period). Patients were included if they completed the regimen as shown by transitioning to maintenance pem, pembro, or pem/pembro. Primary endpoint was CB4. Secondary endpoint was overall survival. Multivariable logistic regression analyses were performed to identify independent predictors of receiving CB4. Results: Among 2090 patients treated during the study period, 966 patients (46%) started on carbo/pem/pembro and subsequently proceeded to maintenance therapy. Among them, 134 patients (14%) received CB4. Multivariable analysis revealed the following statistically significant predictors of receiving CB4, along with respective odds ratios (95% CI) and p-values: Non-white race 1.67 (1.09-2.57) (p=0.017), ECOG ≥1 1.66 (1.07-2.59) (p= 0.024), pandemic year 2020 1.68 (1.11- 2.52) (p=0.013), and ever-smoking status 0.51 (0.30-0.87) (p=0.013). Survival analysis is ongoing. Conclusions: A significant number of patients received CB4. Since this practice is not considered a standard practice, the benefits and risks of CB4 is unknown. Predictors of receiving CB4 included year of treatment, race, ECOG, and smoking status. Further investigation is needed to understand patient-physician decision making process with regards to the length of carboplatin administration.