Background: Early palliative care referral has proven useful in reducing chemotherapy given to cancer patients in the last 30 days of life. We designed a program to help train primary oncologists in having goals of care (GOC) conversations with patients. We conducted this initiative across the Yale Cancer Center/Smilow Cancer Hospital Care Centers (SCHCCs). The aim was to determine whether simple tools could be used by general oncologists to improve patient care, as measured by chemotherapy utilization at the end of life. Methods: A multidisciplinary team completed process mapping and cause-and-effect barrier analysis regarding GOC discussion with patients. Five interventions (PDSA cycles) were chosen to improve physician skills with GOC conversations. SCHCC physicians (n = 21) utilized each tool on 5 patients per month from July to December 2017. Interventions included asking patients simple questions regarding GOC, using the phrase “allow natural death”, giving patients a “serious conversation starter kit”, and identifying patients with poor prognosis. A GOC survey was administered to physicians at baseline and after completion of the initiative. Data on chemotherapy and immunotherapy given during the last 30 days of life for SCHCC patients were measured. Results: SCHCC physicians who completed the baseline survey (n = 17) were 27% likely to discuss GOC repeatedly during the course of illness, compared to 70% post intervention (n = 10), while 88% felt comfortable discussing GOC at baseline compared to 100% after intervention. Chemotherapy within 30 days of death was administered to 30.3% patients (n = 109) in Q1 2017 and 23.3% in Q3 (n = 103) in the Care Centers. Of note, immunotherapy administered within 30 days of death was 6.4% in Q1 and 6.7% in Q3. Conclusions: A quality initiative focused on providing oncologists with simple tools to improve GOC conversations led to decrease from 30.3% to 23.3% of patients receiving chemotherapy within the last 30 days of life. Survey respondents felt more comfortable in discussing GOC and were more likely to do this throughout the course of illness. This pilot suggests that palliative care training of primary oncologists is feasible and may help to decrease unnecessary treatment at the end of life.