Background: Nationally, participation in oncology clinical trials is low, with lack of representation of women and ethnic and racial minorities. Interventions to increase patient enrollment in clinical trials and to ensure equity are necessary to improve cancer care for all patients. Shared decision-making conversations are in physicians’ control and are amenable to intervention. However, we must understand the current communication strategies used by medical oncologists in their discussions with patients about high-stakes treatment decisions, including participation in clinical trials. Methods: This was an observational qualitative study conducted between Jan 2022 to Feb 2023. We recruited US medical oncologists who routinely treat colon cancer and have access to clinical trials. Oncologists participated in a simulated, recorded telehealth encounter with a standardized patient recently diagnosed with advanced colon cancer and referred for discussion of treatment options, including a possible clinical trial. Oncologists were provided with the same background information on the patient. Four actors, a Black female, Black male, White female, and White male, were trained to represent the same patient with the same values. Results: 107 physicians at 42 academic institutions across 27 states were contacted via email, 47 responded (44%). 26 consented, and 21 completed the study. Participants were a median age of 41 (range 31-71 years), 48% female, 29% Asian, 5% Black. Encounter length ranged from 28:25 to 76:12 minutes (average 44:05 min). One to four treatment options were discussed during the encounter, including standard of care (1 or 2 options), clinical trial, and best supportive care. Clinical trials were explicitly discussed in nearly all (20/21) conversations and were introduced as an option 1^st (1/20), 2^nd (5/20), 3^rd (12/20), 4^th (2/20). Clinical trial discussion length ranged from 3:34 to 19:40 min (average 8:30 min). Discussions were shorter with Black vs White patient, average time 7:00 vs 10:00 min, respectively (p = 0.05), female vs male patient was 9:53 vs 7:34 min, respectively (p = 0.12). Recommendations varied from strong recommendation for a clinical trial (9/21), decision deferred to patient (9/21), recommendation for standard of care (3/21). Conclusions: The way oncologists discuss treatment options with their patients powerfully influences the outcome of the conversation and potentially the decision to participate in a clinical trial. In our study, fewer than half of academic GI oncologists recommend a trial. There is significant variability in option talk including the order as well as the content, duration, detail, and recommendations. Clinical trial discussions were shorter for Black patients. A structured approach to this complex decision-making conversation may help oncologists communicate treatment options to patients in a more equitable and effective way.