Scholarly and public debates about criminal appeals have largely taken place in an empirical vacuum. This study builds on our prior empirical work exploring defense-initiated criminal appeals and focuses on criminal appeals by state and federal prosecutors. Exploiting data drawn from a recently released national sample of appeals by state prosecutors decided in 2010, as well as data from all appeals by federal prosecutors to the United States Court of Appeals terminated in the years 2011 through 2016, we provide a detailed snapshot of non-capital, direct appeals by prosecutors, including extensive information on crime type, claims raised, type of defense representation, oral argument and opinion type, as well judicial selection, merits review, and relief. Findings include a rate of success for state prosecutor appeals about four times greater than that for defense appeals (roughly 40% of appeals filed compared to 10%). The likelihood of success for state prosecutor-appellants appeared unrelated to the type of crime, claim, or defense counsel, whether review was mandatory or discretionary, or whether the appellate bench was selected by election rather than appointment. State high courts, unlike intermediate courts, did not decide these appeals under conditions of drastic asymmetry. Of discretionary criminal appeals reviewed on the merits by state high courts, 41% were prosecutor appeals. In federal courts, prosecutors voluntarily dismissed more than half the appeals they filed, but were significantly less likely to withdraw appeals from judgments of acquittal and new trial orders after the verdict than to withdraw appeals challenging other orders. Among appeals decided on the merits, federal prosecutors were significantly more likely to lose when facing a federal defender as an adversary compared to a CJA panel attorney.