Crimes have many features, and the mix of those features can change over time and space. In this article, we introduce the concept of a crime regime to provide some theoretical leverage on collections of crime features and how the collections of features can change. Key tools include the use of principal components analysis to determine the dimensions of crime regimes, visualization methods to help reveal the role of time, summary statistics to quantify crime regime patterns, and permutation procedures to examine the role of chance. Our approach is used to analyze temporal and spatial crime patterns for the city of Los Angeles during an 8-year period. We focus on the number of violent crimes over time and their potential lethality.