In this paper, we assess the impact of recent prison releases on regional crime rates using state level data for the time period 1977 to 1999. We find that a net increase in the population of former inmates positively and significantly affects nearly all felony offenses. The magnitude of our estimates suggest that in a typical year during the 1990s, the net increase in the population of ex-offenders caused by prison releases over the previous year accounts for approximately 2 percent of property crime and 2.5 percent of violent crime. For the violent crimes of murder and robbery, recent prison releases account for substantially larger proportions. However, our results indicate that prisoners released during the 1990s have much smaller effects on crime than prisoners released during earlier time periods. In addition, we find that in states with strong centralized parole boards, prison releases have considerably smaller effects on most crimes, although this pattern is not uniform for all offenses. In conjunction, these two results suggest that states could probably reduce their steady-state prison population from current levels in a targeted and deliberate fashion without appreciably impacting crime rates.