The rapid rise in the nation’s incarceration rate over the past decade has raised questions about how to successfully reintegrate a growing number of exoffenders. Employment has been shown to be an important factor in reintegration, especially for men over the age of 27 that characterize the majority of individuals released from prison. At the same time, there is substantial evidence that employers discriminate against exprisoners. One policy response that has received considerable attention is to deny employers access to criminal history record information, including movements to "ban the box" asking about criminal history information on job applications. The assumption underlying this movement is that knowledge of exoffender status leads directly to a refusal to hire. An alternative view is that some employers care about the characteristics of the criminal history record, and use information about criminal history in a more nuanced, nondiscrete way. This paper explores this question using unique establishment-level data collected in Los Angeles in 2001. On average, we replicate the now common finding that employer initiated criminal background checks is negatively related to the hiring of exoffenders. However, this negative effect is less than complete. The effect is strongly negative for those employers that are legally required to check. But some employers appear to check to gain additional information about exoffenders (and thus hire more exoffenders than other employers), while checking appears to have no affect on hiring exoffenders for those employers not legally required to check. Therefore, initiatives aimed at restricting background checks for those firms not legally required to check may not have the desired consequences of increasing exoffender employment.