BackgroundDespite rising rates of vaccination, quarantine remains critical to control SARS-CoV-2 transmission. COVID-19 quarantine length around the world varies in part due to the limited amount of empirical data.ObjectiveTo assess post-quarantine transmission risk for various quarantine lengths.DesignCohort study.SettingFour US universities, September 2020 to February 2021.Participants3,641 students and staff were identified as close contacts to SARS-CoV-2-positive individuals. They entered strict or non-strict quarantine and were tested on average twice per week for SARS-CoV-2. Strict quarantine included designated housing with a private room, private bathroom and meal delivery. Non-strict quarantine potentially included interactions with household members.MeasurementsDates of exposure and last negative and first positive tests during quarantine.ResultsOf the 418 quarantined individuals who eventually converted to positive, 11%, 4.2%, and 1.2% were negative and asymptomatic on days 7, 10 and 14, respectively. The US CDC recently shortened its quarantine guidance from 14 to 7 days based on estimates of 2.3-8.6% post-quarantine transmission risk at day 7, significantly below the 11% risk we report here. Notably, 6% of individuals tested positive after day 7 in strict quarantine, versus 14% in non-strict quarantine. Ongoing exposure during quarantine likely explains the higher rate of COVID-19 in non-strict quarantine.LimitationsQuarantine should be longer for individuals using antigen testing, given antigen testing’s lower sensitivity than qPCR. Results apply in settings in which SAR-CoV-2 variants do not affect latent period.ConclusionsTo maintain the 5% transmission risk that the CDC used in its guidance, our data suggest that quarantine with qPCR testing 1 day before intended release should extend to 10 days for non-strict quarantine.Funding SourceNone.