Two experiments tested (a) whether jurors make assumptions about the potential punishment that would follow from a guilty verdict, (b) whether those assumptions influence jurors’ implicit threshold for reasonable doubt, and (c) whether informing jurors of the potential punishment additionally influences their implicit threshold. Experiment 1 manipulated the alleged crime (grand theft vs. manslaughter) holding all other factors constant, and found that mock jurors (n = 102, recruited via Amazon Mechanical Turk) had different expectations about the relative punishments but that these expectations did not affect their implicit threshold for reasonable doubt. Experiment 2 manipulated the alleged crime as well as a judicial description of the potential punishment (e.g., term of incarceration of 2–6 vs. 7–20 years). Again, mock jurors (n = 297) were sensitive to the relative punishments, yet their implicit threshold did not differ on this basis. These findings call into question whether jurors should be informed of the potential punishment if the defendant is convicted, an argument advanced by several legal scholars.