The current study is an exploratory application of a technique for modeling developmental trajectories as an illustration of how researchers might use trajectory analysis and county-level data to analyze the effects of imprisonment patterns on crime. Using data from 61 New York counties, the authors model the patterns of imprisonment, defined as the percentage of felons sentenced to state prison in a given year, displayed by the counties from 1990 to 1999, and then demonstrate how knowledge of these patterns can be used to estimate the relationship between imprisonment patterns during the 1990s and crime in 2000. Six distinct patterns of imprisonment are identified, and bivariate and multivariate analyses reveal that high or consistently increasing levels of imprisonment were not associated with lower property or violent crime rates in 2000. Even after controlling for crime in the 1990s and other relevant factors, counties with high or consistently increasing imprisonment had significantly higher violent crime in 2000. Implications for future research are discussed. Figures, tables, appendix, notes, and references (Published Abstract)