Predictors found to be statistically significant of pretrial risk for both Federal and State criminal defendants were: (1) whether there were other charges pending against the defendant at the time of arrest; (2) the number of prior misdemeanor arrests; (3) the number of prior felony arrests; (4) the number of prior failures to appear; (5) whether the defendant was employed at the time of the arrest; (6) the defendant's residency status; (7) whether the defendant suffered from substance abuse problems; (8) the nature of the primary charge; and (9) whether the primary charge was a misdemeanor or a felony. These predictors should be utilized in developing a standardized empirically-based risk assessment instrument to be used for Federal pretrial services. A standardized instrument assists in reducing the disparity in risk assessment practices, provides a foundation for evidence-based practices relating to release and detention, the administration of alternatives to detention programs, and allows for the development of policy that provides guidance to pretrial services agencies regarding release and detention recommendations. This study sponsored by the Office of Federal Detention Trustee (OFDT), tasked to manage and regulate the Federal detention programs, sought to identify statistically significant and policy relevant predictors of pretrial outcome to identify Federal criminal defendants who are most suited for pretrial release and to develop recommendations for the use of OFDT funding that supports the Federal judiciary's alternatives-to-detention program. The study employed data provided by the Administrative Office of the U.S. Courts, Office of Probation and Pretrial Services that described all persons charged with criminal offenses in the Federal courts between October 2001 and September 2007. Figures and appendix