A 1987 survey of Michigan judges helped identify perceived weaknesses in standard criminal jury instructions. Once this survey cited potentially problematic jury instructions, the project developed a short questionnaire that consisted of both true-false and open-ended questions. Actual jurors received this questionnaire immediately after they finished serving on a trial. The project assessed juror comprehension of instructions by comparing the responses of jurors instructed on a particular standard with jurors who had not received such instruction. Of 882 jurors asked to complete the questionnaire, 600 did so for a compliance rate of about 68 percent. Project findings suggested that jury instructions were often lost on jurors and sometimes even backfired. Specifically, the results demonstrated a relatively low rate of comprehension for some concepts among both more and less educated jurors, the apparent ineffectiveness of instructions to improve comprehension, and the negative effect of certain instructions. Nonetheless, some jury instructions did raise comprehension; more educated jurors were generally more likely to grasp the concepts conveyed by an instruction. The project questionnaire is appended. 65 footnotes