COVID-19 (coronavirus disease 2019) is associated with heightened risks of venous and arterial thrombosis and hospitalization due to respiratory failure. To assess whether prophylactic anticoagulation can safely reduce the frequency of venous and arterial thrombosis, hospitalization, and death in nonhospitalized patients with symptomatic COVID-19 and at least one thrombosis risk factor, we conducted the PREVENT-HD double-blind, placebo-controlled randomized trial (A Study of Rivaroxaban to Reduce the Risk of Major Venous and Arterial Thrombotic Events, Hospitalization and Death in Medically Ill Outpatients With Acute, Symptomatic COVID-19] Infection). PREVENT-HD was conducted between August 2020 and April 2022 at 14 US integrated health care delivery networks. A virtual trial design used remote informed consent and clinical monitoring and facilitated data collection through electronic health record integration with a cloud-based research platform. Nonhospitalized patients with symptomatic COVID-19 and at least one thrombosis risk factor were enrolled and randomly assigned to either 10 mg of oral rivaroxaban or placebo daily for 35 days. The primary efficacy outcome was time to first occurrence of a composite of symptomatic venous thromboembolism, myocardial infarction, ischemic stroke, acute limb ischemia, non-central nervous system systemic arterial embolism, hospitalization, or death through day 35. The principal safety end point was International Society on Thrombosis and Hemostasis critical-site or fatal bleeding. The last study visit was on day 49. The study was terminated prematurely because of enrollment challenges and a lower-than-expected blinded pooled event rate. A total of 1284 patients underwent randomization with complete accrual of primary events through May 2022. No patients were lost to follow-up. The primary efficacy outcome occurred in 22 of 641 in the rivaroxaban group and 19 of 643 in the placebo group (3.4% versus 3.0%; hazard ratio, 1.16 [95% CI, 0.63-2.15]; P=0.63). No patient in either group experienced critical-site or fatal bleeding. One patient receiving rivaroxaban had a major bleed. The study was terminated prematurely after enrollment of 32% of planned accrual because of recruitment challenges and lower-than-expected event rate. Rivaroxaban prescribed for 35 days in nonhospitalized patients with symptomatic COVID-19 at risk for thrombosis did not appear to reduce a composite end point of venous and arterial thrombotic events, hospitalization, and death. URL: https://www. gov; Unique identifier: NCT04508023.