Abstract: Dynamic origin-destination(OD) matrix is an important basis for the dynamic operation management of urban rail transit, so it is significant to accurately estimate the dynamic OD matrix for the improvement of the actual operation management level of urban rail transit. This article analyzed the influencing factors of the temporal and spatial distribution for the OD passenger flow, such as the space influencing factors including land-use type, attracted traffic flow, line properties, if the starting point and the terminal were on the same line, and the time influencing factors including riding time, transfer times and distance. This article built the estimation model of passenger flow dynamic OD matrix for the urban rail transit, selected the local area of Beijing urban rail transit network as a case study, analyzed the estimation effect of the model. Comparing the proposed model with historical data, the estimation accuracy of the morning and evening peak was improved by 4% ~ 10%, the flat peak was improved by 8% ~17%. Meanwhile, under the granularity of 15 min, 30 mins, and 60 mins, the full-day average accuracy was increased by 8.67%, 11.75%, 3.46%.