This paper presents a new theory of urban pedestrian movement, arguing that the configuration of the urban grid is the primary generator of movement patterns, rather than the traditional 'attraction' theory which focuses on land uses as attractors. The paper introduces the concept of 'natural movement' as the proportion of pedestrian movement determined by the grid configuration itself. It is suggested that retail land uses act as multipliers on this basic pattern, and that the 'space syntax' measure of 'integration' is the most important configurational correlate of movement patterns. The paper argues that urban design should focus on configurational relations to the larger urban system rather than local properties of a space. The paper discusses the effects of configuration on both through-movement and to-movement in urban grids, which are independent of built-form attractors. It presents case studies, including King's Cross, to illustrate how configuration influences movement patterns. The paper also discusses the relationship between spatial configuration and movement, showing that integration values are strongly correlated with movement rates. The paper argues that the relationship between integration and movement is logarithmic, suggesting that shops act as logarithmic multipliers on the basic pattern of movement defined by configuration. The paper concludes that the distribution of movement rates at the area level is itself logarithmic, reflecting both global properties of the grid and differences in the distribution of attractors. The paper also discusses the effects of 'negative attractors' and suggests that the low movement rates in housing estates are a configurational effect rather than an attractor effect. The paper argues that the almost uniform blight surrounding the King's Cross site is a long-term configurational effect.This paper presents a new theory of urban pedestrian movement, arguing that the configuration of the urban grid is the primary generator of movement patterns, rather than the traditional 'attraction' theory which focuses on land uses as attractors. The paper introduces the concept of 'natural movement' as the proportion of pedestrian movement determined by the grid configuration itself. It is suggested that retail land uses act as multipliers on this basic pattern, and that the 'space syntax' measure of 'integration' is the most important configurational correlate of movement patterns. The paper argues that urban design should focus on configurational relations to the larger urban system rather than local properties of a space. The paper discusses the effects of configuration on both through-movement and to-movement in urban grids, which are independent of built-form attractors. It presents case studies, including King's Cross, to illustrate how configuration influences movement patterns. The paper also discusses the relationship between spatial configuration and movement, showing that integration values are strongly correlated with movement rates. The paper argues that the relationship between integration and movement is logarithmic, suggesting that shops act as logarithmic multipliers on the basic pattern of movement defined by configuration. The paper concludes that the distribution of movement rates at the area level is itself logarithmic, reflecting both global properties of the grid and differences in the distribution of attractors. The paper also discusses the effects of 'negative attractors' and suggests that the low movement rates in housing estates are a configurational effect rather than an attractor effect. The paper argues that the almost uniform blight surrounding the King's Cross site is a long-term configurational effect.