In an effort to examine the interaction of judicial controls and the defendant's character on jury decisions, both the level of biasing information and the timing of specific judicial instructions designed to control bias were varied in the account of a closely contested burglary trial. Judicial Controls were specific judicial instructions on the presumption of innocence, the nature of reasonable doubt, and the guidelines of what should and should not be considered courtroom evidence. Subjects from two community college samples and one sample of actual jurors were employed. The trial transcript was presented with the addition of irrelevant but damaging trial information to one-half of the subjects and without such information to the other half. Accordingly, one-half of each group received pretrial judicial controls while the other half of each group did not. Subjects were asked to read the trial account and then to render a verdict of guilty or not guilty, to indicate their rating of the strength of evidence against the defendant, and to record the time required for deliberation. While student jurors were not affected, actual jurors differentially responded to the instruction variable convicting the defendant significantly more often in the absence of pretrial judicial controls. For all subjects, regardless of either sample membership or assigned treatment condition, guilty verdicts corresponded to the jurors' view of the trial evidence as more incriminating to the defendant. Moreover, comparisons between samples revealed that actual jurors voted guilty more often, gave higher strength of evidence ratings, and deliberated longer than student jurors. The only exceptions were found among jurors given pretrial judicial controls (in terms of verdicts) and among 'biased' jurors (in terms of strength of evidence ratings). Perhaps the most significant finding was the notable absence of any effects of biasing information. Tabular data, 116 references, pilot study results, the verdict forms, survey instruments, and the trial transcript are appended. (Author abstract modified).