Background: To improve value-based cancer care, we designed an intervention using lay health workers (LHWs) who we trained to assess patient's symptoms. LHWs referred patients to palliative care and/or behavioral health in response to positive assessments. We implemented the intervention in collaboration with a community oncology group and a Medicare Advantage payer for patients with all stages of cancer. At ASCO 2019 we demonstrated the associated reductions in patient-reported symptoms, acute care use, and total costs of care. This study evaluates the effect of the intervention on end-of-life cancer care, specifically healthcare use in the last month of life and total costs of care from diagnosis until death. Methods: We enrolled all newly diagnosed health plan beneficiaries with solid and hematologic malignancies from 11/2016 through 9/2018 and compared outcomes to all patients diagnosed with cancer in the year prior to the intervention (control arm). For all patients who died within 12-months follow-up, we compared risk of death using Cox Models and generalized linear regression to compare healthcare use in the last month of life, and total costs of care from diagnosis until death. All models were adjusted for age, stage, comorbidities, diagnosis, and length of follow-up. Results: 180 patients in the intervention and 156 in the control died during the study. In both groups, the mean age was 80 years; 49% were non-Hispanic White, 40% Hispanic, 5% Asian/Pacific Islander, and 4% black. There were no differences in survival (HR 0.96, p = 0.6). Intervention patients as compared to the control had lower mean inpatient admissions (0.3 +/- 0.04 versus 0.5 +/- 0.07, p = 0.02) and emergency department visits per thousand members per year (0.11 +/- 0.02 versus 0.49 +/- 0.06, p < 0.001) in the last month of life, greater proportion of patients with hospice use (69% versus 48%, p < 0.001), lower proportion of patients with acute care facility deaths (37.5% versus 62.5%, p = 0.02) and lower median total healthcare costs from diagnosis until death ($24,902 versus $33,145, p = 0.02). Conclusions: An LHW intervention significantly improved the value of end-of-life cancer care and may be a solution to improve burdensome and costly care for patients.