Objectives. To examine shelter characteristics and infection prevention practices in relation to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection point prevalence during universal testing at homeless shelters in the United States.Methods. SARS-CoV-2 testing was offered to clients and staff at homeless shelters, irrespective of symptoms. Site assessments were conducted from March 30 to June 1, 2020, to collect information on shelter characteristics and infection prevention practices. We assessed the association between SARS-CoV-2 infection prevalence and shelter characteristics, including 20 infection prevention practices by using crude risk ratios (RRs) and exact unconditional 95% confidence intervals (CIs).Results. Site assessments and SARS-CoV-2 testing results were reported for 63 homeless shelters in 7 US urban areas. Median infection prevalence was 2.9% (range = 0%-71.4%). Shelters implementing head-to-toe sleeping and excluding symptomatic staff from working were less likely to have high infection prevalence (RR = 0.5; 95% CI = 0.3, 0.8; and RR = 0.5; 95% CI = 0.4, 0.6; respectively); shelters with medical services available were less likely to have very high infection prevalence (RR = 0.5; 95% CI = 0.2, 1.0).Conclusions. Sleeping arrangements and staffing policies are modifiable factors that might be associated with SARS-CoV-2 infection prevalence in homeless shelters. Shelters should follow recommended practices to reduce the risk of SARS-CoV-2 transmission.