Two jury simulation experiments tested participants’ sensitivity to variations in the probative value of a piece of negative forensic evidence: failure to find gunshot residue (GSR) on a defendant alleged to have fired a gun. Experiment 1 found that if no GSR was detected, juries (N = 115) of undergraduates were appropriately less likely to convict a criminal defendant when the probability of detecting GSR was high than when it was low. Participants were unaffected by contextualizing expert testimony that emphasized either the value of the negative evidence for making inductive inferences or that the negative evidence was inconclusive for making deductions. Experiment 2 used a sample of venire jurors (N = 420) and manipulated the probability of detecting GSR (0 percent, 50 percent, 60 percent, 90 percent, or 100 percent) given that a gun was fired. Consistent with the first experiment, venire jurors were more likely to convict when the probability of detection was 0 percent or 50 percent than when it was 100 percent, but verdicts did not differ between the middle groups. This pattern of results suggests that jurors may evaluate negative evidence according to a fairly crude metric—giving it no weight if the probability of detection is zero, a great deal of weight if the probability of detection is 100 percent, and moderate weight if the probability of detection is somewhere in between.