Smart card fare collection systems are widely used nowadays in urban public transport networks. These systems are bound to facilitate the collection and management of revenues in transit authorities. However, since smart card systems collect a large amount of data on a daily basis, they can be exploited to better characterize the demand and supply of public transport in subways, tramways and bus networks, while data at an individual level should remain strictly confidential. The spatial and temporal dimensions of the data make it very interesting for planning purposes, but the data must first be validated and completed before further analysis. This article presents the results of five years of research conducted in collaboration with theSociété de transport de l'Outaouais, in Quebec. The following analyses are presented: error processing, estimation of alighting points, diffusion of operational statistics, analysis of user behaviour, analysis of network performance, comparison with household survey data and user loyalty modelling. © 2011 INRETS et Springer-Verlag France.