This study examined how the characteristics of people with mental illness who are participants in post-booking jail diversion programs affect recidivism and time spent incarcerated. The study employed data from a multi-site, federally funded jail diversion initiative. A pre-post comparison design was used to compare experiences of arrest and days spent in jail of diverted individuals for the 12 months following enrollment with the 12 months prior to enrollment. Also compared were differences in 12-month public safety outcomes. Data were collected between February 2003 and August 2007. The findings suggest that people with mental illness who are diverted from jail to community-based services experience fewer arrests and jail days. Approximately half of the sample were never arrested during the 12 months following enrollment. The strongest differences between people who experienced reduced contact with the criminal justice system and people with unchanged or increased contact were found in measures of criminal history. The results suggest that services targeted to diverted individuals with mental illness should address public safety goals, not just those of public health.