A primary goal of this research was to empirically evaluate a set of assumptions, advanced in the Supreme Court's ruling in Buchanan v. Angelone (1998), about jury comprehension of death penalty instructions. Further, this research examined the use of evidence in capital punishment decision making by exploring underlying mediating factors upon which death penalty decisions may be based. Manipulated variables included the type of instructions and several variations of evidence. Study 1 was a paper and pencil study of 245 undergraduate mock jurors. The experimental design was an incomplete 4×2×2×2×2 factorial model resulting in 56 possible conditions. Manipulations included four different types of instructions, presence of a list of case-specific mitigators to accompany the instructions, and three variations in the case facts: age of the defendant, bad prior record, and defendant history of emotional abuse. Study 2 was a fully-crossed 2×2×2×2×2 experiment with four deliberating mock juries per cell. Manipulations included jury instructions (original or revised), presence of a list of case-specific mitigators, defendant history of emotional abuse, bad prior record, and heinousness of the crime. The sample of 735 jury-eligible participants included 130 individuals who identified themselves as students. Participants watched one of 32 stimulus videotapes based on a replication of a capital sentencing hearing. The present findings support previous research showing low comprehension of capital penalty instructions. Further, we found that higher instruction comprehension was associated with higher likelihood of issuing life sentence decisions. The importance of instruction comprehension is emphasized in a social cognitive model of jury decision making at the sentencing phase of capital cases.