For this analysis, a method was developed for measuring the propensities of would-be jurors to convict. Propensity refers to a personal value judgment or normative criterion rather than an empirical or statistical pattern. The measurement system has been applied on a continuing basis, using a roughly random sample of students at the University of Illinois, as part of a series of exercises performed by graduate students and law students enrolled in a seminar. The basic decision theory model refers to a perspective for analying decisions which emphasize the fact that decisionmakers choose the decision which they implicitly perceive as providing the greatest expected benefits minus costs of all the alternative decisions available. Questionnaires were employed to determine the threshold probabilities of various types of respondents in rape, robbery, and consumer fraud cases. Findings revealed that jurors in criminal cases apply a standard of guilt that is substantially below the meaning of 'beyond a reasonable doubt.' In consumer fraud cases, business students indicated higher threshold probabilities than liberal arts students. In rape cases, the conviction threshold probabilities of males are substantially higher than those of females. Jury instructions would be more helpful to jurors if they included a statement about determination of the probability of the defendant's guilt. Few jurisdictions provide instructions for jurors as to the meaning of 'beyond a reasonable doubt,' and thus the juror's propensity to convict or acquit assumes greater importance in the decision. One table, 3 figures, 27 footnotes, and approximately 50 references are included; a glossary of symbols and terms is appended.