This paper investigates the spatial aspect of criminal activity in Vancouver, Canada, employing social disorganization theory, routine activity theory and multiple measures of crime. Crime counts and crime rates with residential and ambient populations as denominators are calculated using the calls for service made to the Vancouver Police Department. The ambient population-a 24-hour average estimate of a population in a spatial unit to capture the population at risk-is obtained from the LandScan Global Population Database and calculated at a spatial resolution relevant to criminological research. Utilizing a spatial regression technique, strong support is found, for routine activity theory across space and the use of ambient populations -when calculating crime rates and measuring the population at risk.