To predict juror partiality, surveys were made of two systematically drawn samples of the jury roster in Yolo County. A total of 323 persons were interviewed through the use of a 45-item schedule comprised largely of forced-choice questions. A similar 57-item schedule was used to interview 369 persons. The surveys focused upon respondent prejudgment regarding three criminal trials due to be held there -- one of which was subsequently moved to another county -- and the correlates and causes of such prejudgment. Respondents indicating partiality in a given case by believing the defendant to be guilty or by feeling incapable of being impartial were compared to those evincing no such partiality. The comparison was made using three sets of variables putatively associated with such partiality. The findings showed a strong statistical relationship in each of the cases between respondent scores on an information index devised for that case and both dependent variables, particularly the measure of prejudging propensity involving belief in the defendant's guilt. In addition, there was a strong connection between a respondent's level of information about a specific case as measured by the information indexes and media usage. For example, those attentive to both newspapers and television news proved more highly informed about the cases than did single media attenders, and they, in turn, were more highly informed than those who attended to neither medium. General attitudes toward crime and punishment issues were also probed through the use of an attitude-on-crime index. Respondents holding conservative attitudes were more likely than liberal respondents to hold prejudging opinions. The respondent's social background or demographic characteristics were also considered and strong relationships to prejudgment were found. Discriminant function analysis used to assess the relative strength of each of the variables showed that the information index accounted for the largest part of the variation in prejudgment. Statistical data and about 40 references are included.