This paper develops a test for identifying racial bias in bail decisions. The analysis is motivated using Becker’s model of racial bias, which posits that if bail judges are racially unbiased, rates of pretrial misconduct will be the same for marginal white and marginal black defendants. If bail judges are racially biased, marginal white defendants will have higher rates of misconduct than marginal black defendants. This model is tested using the release tendencies of bail judges, that are quasi-randomly assigned, to identify the relevant race-specific misconduct rates. Approximations from Miami and Philadelphia demonstrate that bail judges are racially biased against black defendants and that racial bias is substantially higher with both inexperienced and part-time judges.