A total of 60 male and 60 female introductory psychology students were randomly assigned to 4 conditions of evidence withholding or to a fifth condition which served as a control condition for unambiguous unit formation between the defendant and the crime. Each subject received a booklet in which the crime was described and trial testimony was recorded. The manipulation of evidence withholding was accomplished by varying the content of the defendant's response to the prosecutor's final question. For example, the defendant in the strong withholding condition responded by taking the fifth amendment and therefore declining to answer. Following the testimony, subjects were given a juror assessment form, which consisted of the dependent variables and a check on the manipulation of evidence withholding. Analysis of subjects' responses to questions of innocence, recommended verdict, and perceived degree of likelihood that the defendant attempted to withhold evidence from the injury confirmed the hypothesis. Study results suggest that since most jurors perceive a fifth amendment plea as a veritable admission of guilt, other means of protection against self-incrimination that do not lead to a presumption of guilt by their invocation should be devised. Footnotes, a table, and 10 references are included.