Background: Previously, among Veterans with cancer, lay health workers (LHWs) trained to discuss patients’ goals of care reduced acute care use, improved patient experiences and reduced total costs of care at the end-of-life. Among Medicare-Advantage beneficiaries with cancer, LHWs trained to proactively assess patient symptoms reduced symptom burden, acute care use and total costs of care. It is unknown whether LHWs can assist with both goals of care and symptom assessments in community settings. The objective of this randomized clinical trial was to determine the effect of a LHW-led goals of care and symptom assessment intervention on acute care use and secondarily goals of care documentation, satisfaction and end-of-life healthcare use among patients with advanced cancer in a community practice. Methods: Newly diagnosed patients with advanced stages of solid and hematologic malignancies who planned to receive care at the oncology practice were randomized from 8/11/2016 through 2/5/2020 into intervention and control groups. Patients completed validated satisfaction surveys at randomization and 9 months follow-up and were followed for 12 months. We compared risk of death using Cox Models, healthcare use and satisfaction using generalized regression models adjusted for length of follow-up. Results: 128 patients were randomized; 64 in the intervention and 64 in the control. The mean age was 67 years; 22% identified as Hispanic/Latino; 57% White, 30% Asian Pacific Islander, 8% Black or African American, 1% Native Hawaiian, 1% American Indian/Alaskan Native, 3% multiple races/ethnicities. There were no survival differences. Intervention patients were less likely to utilize the emergency department (OR: 0.35; 95% CI 0.17-0.72) and hospital (OR: 0.48; 95% CI 0.23-0.98) and had lower mean emergency department visits (1.05 +/- 1.74 versus 1.84 +/- 2.55, p = 0.04) and hospitalizations per year (0.63 +/- 1.28 versus 1.26 +/- 2.23, p = 0.04) as compared to control patients. More intervention patients had their goals of care documented (94% versus 52% p < 0.001) and used hospice (35% versus 14% p = 0.004) as compared to control patients. There were no differences in palliative care use (89% versus 77% p = 0.09). At 9 months follow-up as compared to baseline, intervention patients experienced greater improvements in satisfaction with care (difference-in-difference: 0.41, 95% CI 0.22-0.60, p < 0.001). Among 30 patients who died (n = 16 intervention; n = 16 control), more patients in the intervention used hospice (81% versus 43%) and fewer used acute care in the last month (37% versus 81%, p = 0.012) than in the control. Conclusions: An LHW intervention reduced acute care use among patients with cancer, improved patient experiences and end-of-life care. This intervention may be a scalable approach to improve care delivery and experiences for patients after a diagnosis of cancer. Clinical trial information: NCT03154190