Thousands of criminal files were analyzed by research assistants to determine which characteristics of the case and the offender, as documented in the files, were most closely related to the various judicial decisions. Studies were also conducted at local courts where direct observations were made of the sentencing consistency based upon the determined culpability of the offender. The volume of cases, however, prevents the sentencer from making a detailed examination of the offender's culpability, so that crude and objective indicators of culpability (i.e., the number of offenses, their joint financial consequences, and the number of former convictions). It is only with unique cases that the sentencer appears to depart from routine considerations in determining the sentence. Although each sentencer generally aims at personal consistency and the objective criteria used by all are basically the same, there is a consistent disparity in sentencing among the sentencers. This finding suggests the necessity of sentencing guidelines, which would prompt all sentencers to seek consistency within the same guidelines rather than in their own personal sentencing histories. Eleven references are provided.