Urban railways in Japan such as those in the capital region are subject to chronic train delays, which have led to a decline in punctuality and quality of service. Therefore, to reduce these delays, it is necessary to clarify their mechanisms. Previous studies attempted to clarify the mechanisms of delays in urban railways. Rainfall is one of the causes of delays as it affects congestion factors due to concentration of passengers and vehicle performance such as acceleration and braking, but few studies have quantitatively analysed and clarified its impact. In addition, as the impact of rain on the train running performance also varies with temperature, the impact is deemed to be seasonal, but few studies have clarified this observation. Therefore, this study aims to quantitatively clarify the impact of rainfall by station and train using train operation record data and weather data from the Tokyo Metro Tozai Line acquired in seconds, and further analyse whether it is seasonal.