This paper explores the influence of the structure and daily rhythms of the city, the repeated ebb and flow of populations undertaking ‘routine activities’, upon the spatio-temporal patterning of crime. Facilitated by access to high quality population (flow) data, as well as spatially and temporally disaggregated crime data, the paper is structured in response to four key questions: (1) Over what spatial and temporal frames are crime patterns meaningful? (2) What is the trade-off between the spatial concentration of crime and crime pattern dynamics? (3) What population-at-risk measure should be used to quantify crime and how does this impact upon the identification of crime hotspots? and (4) Are the characteristics of ambient populations and urban areas associated with crime hotspots? The provocation emergent from these diverse findings centres on the need to advance of causal models of crime sensitive to space and time, whilst also capable of taking advantage of new data.