"This paper presents evidence of the effects of pretrial detention status on criminal case outcomes in federal criminal cases. I find that criminal defendants who are released pending trial earn a roughly 72 percent decrease in sentence length and a 36 percentage-point increase in the probability of receiving a sentence below the recommended federal sentencing Guidelines range. Pretrial release also reduces the probability that a defendant will receive at least the mandatory minimum sentence—when one is charged—by 39 percentage points, but does not affect the probability that the defendant will face a mandatory minimum sentence. To address the identification problem inherent in using pretrial detention status as an explanatory variable, I take advantage of the fact that pretrial release in federal courts is typically determined by magistrate judges who vary in their propensities to release defendants pending trial. This setting allows magistrate judge leniency to serve as an instrumental variable for pretrial release. I also present suggestive evidence of the mechanism at work. It appears that pretrial release affects case outcomes in two distinct ways: most importantly, by giving defendants the opportunity to present mitigating evidence at sentencing and, secondly, by making it easier for defendants to earn a sentencing reduction by providing substantial assistance to the government. In contrast, this paper does not find evidence that pretrial release improves defendants’ abilities to bargain with prosecutors. I also find that the effects of pretrial detention status on case outcomes are heterogeneous, and most pronounced for drug offenders."