The Judicial Conference of the United States passed a resolution in 1964 recommending that district and appeals court judges limit the publication of opinions to those which are of general precedential value. By 1972, the Judicial Conference was asked to have each circuit review its guidelines for publication and focus again on limiting publication further through the modification of existing rules. As of 1978, the Judicial Conference had received enough information to conclude that one opinion publication plan was not preferable over another and that a model plan could not be recommended due to insufficient consensus on legal and policy matters. Thus, each circuit continues to operate under its own criteria for determining whether a decision merits publication. The study of unpublished decisions in three circuits demonstrates that official criteria for publication do not provide an adequate description of the differences between published and unpublished decisions. A significant number of unpublished decisions in appeals courts appear to involve nonconsensual appeals in which judges exercised substantial discretion. Rules governing publication in the three circuits are stated in very broad language, and no objective and precise guidelines are included. Judges are admonished to avoid publishing decisions if their opinions have no precedential value, although the interpretation of such rules is subjective. 5 tables.