The concept of the night-time economy has gained attention in recent years as cities around the world recognize the potential of night-time activities. Most research on the night-time economy adopts a social perspective, without examining this topic from an economic approach, probably because official statistics do not provide suitable information for this purpose. The availability of data on bank card transactions opens new avenues in the study of the night-time economy by providing information on night-time spending in city establishments. This article explores the spatial and temporal patterns of night-time spending by combining different data sources -bank transaction data, mobile phone data, social networks, cadastral records, etc.-, taking the city of Madrid as the study area. The results reveal that night-time spending in establishments in the city of Madrid is equivalent to approximately 30 % of total spending. They also show that the spatial concentration of night-time spending increases as the night progresses, and that the structure of spending is much more diversified in the evening than at dinner time and late at night. Explanatory models (Ordinary Least Squares and Random Forest regression) show that variables related to bars and restaurants, retail area, popularity of establishments, and tourist accommodations have strong explanatory power over the spatial distribution of night-time spending. The article shows that the analysis of bank card transaction data provides information of great interest to urban managers and planners.