Background: Documenting GOC is integral to patient care and quality performance but has been underutilized by oncologists due to many barriers. As oncologists play a key role in initiating GOC discussions, we implemented a clinical initiative to improve their GOC documentation and evaluated the impact of such documentation on patient care during the EOL (last 30 days of life). Methods: We launched the initiative among 270 medical oncologists in an academic cancer center in 4/2020. A newly formulated GOC note to ease documentation was embedded in oncology outpatient and inpatient notes. Oncologists completed components in the GOC note that applied to their communication about GOC with the patient: 1) cancer natural history, 2) patient goals, and 3) EOL discussion: patient resuscitation preferences and, when pertinent, receptivity to hospice referral. GOC notes were pulled to a centralized location in the electronic health record (EHR) that displays documents relevant to patients’ values, goals and preferences. A dashboard allowed continual monitoring of documentation performance. We evaluated the association between GOC notes and outcomes of patient care at EOL. We further analyzed the impact of EOL discussion on EOL care. Results: The GOC note completion rate steadily rose after implementation. GOC notes were present in EHR for 46% of 10,006 patients who were either seen in outpatient clinic or discharged from hospital during the 1^st week of January 2021. Among 1790 patients who died between 7/1/20 and 12/31/20 and had either at least an outpatient visit or hospitalization during EOL, the median days from first GOC note and first EOL discussion to the patient’s death were 71 days and 24 days, respectively. Linear regression analysis demonstrated that patients who had GOC note 60 days before death spent less time as inpatient during EOL (0.4 day less/patient, from 8.1 to 7.7, P = 0.01). When EOL discussion was documented 30 days before death, patients also spent less time in the hospital (1.2 days less/patient, from 9.7 to 8.5, P < 0.001) and in the ICU (0.3 days less/patient, from 1.7 to 1.4 ICU days, P = 0.04), and were 4% less likely to receive chemotherapy (from 38% to 34%, P = 0.004) at EOL. During the same period, among 1,009 patients with hospital admission in the last 30 days of life, those with a prior documented EOL discussion had shorter inpatient stay (7.7 vs 13.1 days, P < 0.001) and were more likely to be discharged to hospice (34% vs 22%, P = 0.003). Conclusions: During the COVID-19 pandemic, we successfully implemented GOC documentation by medical oncologists that is easily visible by the full care team. Documentation of GOC including EOL discussion was associated with fewer days in the hospital and ICU, increased hospice referral, and lower likelihood of receiving chemotherapy during patients’ last 30 days of life.