This paper investigates why crime rates are higher in cities than in small cities or rural areas. Using data from victimization surveys, the National Longitudinal Survey of Youth (NLSY), and the Uniform Crime Reports (UCR), the authors explore the factors contributing to the urban crime premium. They decompose the urban crime effect into four categories: (1) higher pecuniary returns to crime in urban areas, (2) lower probability of arrest or recognition in urban areas, (3) characteristics that affect crime and are exogenous to location but correlated with urban status, and (4) characteristics that are endogenous to location and both cause and are caused by urban status. The authors find that approximately 45-60% of the urban crime premium can be explained by observable characteristics of individuals and cities. These characteristics include tastes, social influences, and family structure. The remaining 27% of the effect is attributed to higher pecuniary returns to crime in cities, while the remaining 20% is due to lower arrest probabilities and recognition. The authors also find that the urban crime premium is influenced by the characteristics of individuals and cities, such as the percentage of female-headed households, which can dominate the effect when included in regressions. The paper also discusses the role of urban anonymity in reducing the probability of recognition and reporting of crimes. However, the effect of anonymity is limited, as it may not fully explain the urban crime premium. The authors conclude that while some factors such as higher pecuniary returns and lower arrest probabilities contribute to the urban crime premium, the majority of the effect is due to characteristics of individuals and cities that are influenced by urban living. These characteristics include social influences, family structure, and tastes, which may lead to higher crime rates in urban areas. The paper highlights the complexity of the relationship between cities and crime and the need for further research to understand the underlying mechanisms.This paper investigates why crime rates are higher in cities than in small cities or rural areas. Using data from victimization surveys, the National Longitudinal Survey of Youth (NLSY), and the Uniform Crime Reports (UCR), the authors explore the factors contributing to the urban crime premium. They decompose the urban crime effect into four categories: (1) higher pecuniary returns to crime in urban areas, (2) lower probability of arrest or recognition in urban areas, (3) characteristics that affect crime and are exogenous to location but correlated with urban status, and (4) characteristics that are endogenous to location and both cause and are caused by urban status. The authors find that approximately 45-60% of the urban crime premium can be explained by observable characteristics of individuals and cities. These characteristics include tastes, social influences, and family structure. The remaining 27% of the effect is attributed to higher pecuniary returns to crime in cities, while the remaining 20% is due to lower arrest probabilities and recognition. The authors also find that the urban crime premium is influenced by the characteristics of individuals and cities, such as the percentage of female-headed households, which can dominate the effect when included in regressions. The paper also discusses the role of urban anonymity in reducing the probability of recognition and reporting of crimes. However, the effect of anonymity is limited, as it may not fully explain the urban crime premium. The authors conclude that while some factors such as higher pecuniary returns and lower arrest probabilities contribute to the urban crime premium, the majority of the effect is due to characteristics of individuals and cities that are influenced by urban living. These characteristics include social influences, family structure, and tastes, which may lead to higher crime rates in urban areas. The paper highlights the complexity of the relationship between cities and crime and the need for further research to understand the underlying mechanisms.