To understand whether commuters will take rail transit during the COVID-19 pandemic,a logistic regression model was constructed from three aspects of personal attributes,travel attributes and perception of COVID-19 based on 559 valid questionnaires.The results show that:occupation,commuting tools before the COVID-19 pandemic,walking time from residence to the nearest subway station,the possibility of being infected in private car and the possibility of being infected in public transport have significant influence on the commuters'choice of rail transit.Self-employed people and freelancers,commuters who used non-public transport before the COVID-19 pandemic,and commuters who take longer to walk from their residences to the nearest subway station are less likely to commute by rail transit during the COVID-19 pandemic.Commuters who think that the risk of being infected with the virus in public transport is higher have a lower probability of choosing rail transit.The confidence in bus/subway/taxi/taxi-hailing of commuters who do not choose to commute by rail transit during the COVID-19 pandemic is not high.The study of this paper can provide reference for the formulation of urban rail transit control measures during the COVID-19 pandemic,so as to formulate more perfect measures to ensure the safety of the returning workers.