In judicial behavior literature, the nonunanimous decision has emerged as the primary measure of conflict in appellate courts. This focus can produce invalid inferences about disparate attitudes of judges on courts that have a high unanimous decision rate. For example, judges in some courts traditionally acquiesce to the member who writes the opinion. The potential effects of the panel system upon the distribution of conflict are demonstrated through two hypothetical examples of court of appeals decisionmaking which consider both varying and identical levels of support for decisions. The patterns of different combinations of five judges deciding on criminal cases show that individual positions of judges can vary considerably without any sign of dissent appearing in the final ruling. Thus, the absence of dissent does not necessarily indicate that the judges agree over how certain issues should be treated. This situation is best described as lack of uniformity among individual viewpoints, or dissensus. Statistical measures of central tendency and dispersion are developed to assess the amount of dissensus in unamimous decisions of the court of appeals. An analysis of criminal appeals decisions recorded in the Federal Reporter from 1966 to 1970 revealed a dissent rate of 6.2 percent. A progrovernment score was computed for each appointed member of the court who had heard at least one such case. Means, standard deviations, and coefficients of variation were then calculated from these support scores for each circuit. Examination of these data, as illustrated by tables, demonstrate that appellate judges are by no means in agreement as to how criminal defendants' appeals should be treated. The influence of workloads and en banc sessions on unanimous decisions was also explored. The findings suggest that impediments to the expression of dissent along with the availability of outlets to express disagreements with other court members operate jointly to affect the manifestation of dissensus. Although further research is clearly needed, this paper has proved that valuable data can be gathered from unanimous decisions if appropriate methodologies are devised. A bibliography with eight references is provided. (Author abstract modified)