Integrating routine SARS-CoV-2 testing in dialysis facilities may benefit patients receiving dialysis by mitigating risks of serious illness and reducing transmission. Patient acceptance of nonmandatory testing is unknown. To evaluate the acceptance of 2 SARS-CoV-2 testing strategies among patients in hemodialysis facilities nationwide. This nationwide cluster (dialysis facility-level) randomized trial investigated the acceptance of SARS-CoV-2 testing among patients receiving maintenance hemodialysis at facilities located in 22 states. Anterior nares real-time reverse transcriptase-polymerase chain reaction tests offered once every 2 weeks (static testing facilities) vs offered once a week, once every 2 weeks, or once a month depending on county COVID-19 infection prevalence (dynamic testing facilities). Facilities were randomized by county, and tests were offered for 3 months between February 4 and July 24, 2023. The primary outcome was test acceptance. Secondary outcomes included the proportion of patients who accepted at least 1 test. In total, 62 hemodialysis facilities were randomized and 57 participated. Among 2389 participating patients, the median age was 64 (IQR, 54-74) years, 1341 (56%) were male, 138 (6%) were categorized as American Indian, 60 (3%) Asian, 885 (37%) Black, 75 (3%) Native Hawaiian or Pacific Islander, 338 (14%) Hispanic, and 876 (37%) White; and 1603 (67%) had diabetes. A median of 6 (IQR, 6-6) tests were offered per patient in the static arm and 4 (3-6) tests in the dynamic arm. Test acceptance was low: 8% of offered tests were accepted in each of the test arms. Among 503 patients who accepted at least 1 test, the median percentage of offered tests that were accepted was 16% (IQR, 17%-42%) using the static testing strategy and 50% (IQR, 33%-75%) using the dynamic testing strategy (P < .001). Older patients (odds ratio [OR], 1.08 [95% CI, 1.01-1.16] per 5-year age increment), patients with (vs without) diabetes (OR, 1.59 [95% CI, 1.18-2.16]), and women compared with men (OR, 1.30 [95% CI, 0.98-1.73]) were more likely to accept multiple tests. Patients designated in the electronic health record as Hispanic were more likely than patients designated as White (OR, 1.78 [95% CI, 1.15-2.76]) to accept at least 1 test, whereas patients living in zip codes electing Republican representatives to Congress were less likely than patients living in zip codes electing Democratic representatives (OR, 0.34 [95% CI, 0.17-0.69]) to accept multiple tests. In this cluster randomized trial evaluating 2 SARS-CoV-2 testing strategies in dialysis facilities, test acceptance was low, and a dynamic testing strategy anchored to COVID-19 infection prevalence did not outperform a static testing strategy of every 2 weeks. ClinicalTrials.gov Identifier: NCT05225298.