The analysis considered the degree of agreement between the recommended and actual sentences in 328 individual cases and also used experimental manipulations of 157 cases to determine the influence of different sections of the reports. The 157 reports were each divided into 5 sections: initial information, demographic information, general information presented by the probation officer, summary and conclusions, and the recommendation. The reports were modified by changing the sequence of the five sections and by inserting three decision points. Eight different sequences were provided. The judge reading the report indicated the intended disposition at each of the three decision points. Thirteen judges took part in the study. The judges used all the sections to some extent. They analyzed the recommendation sections in detail and were strongly influenced by them, even though they rated them low in usefulness. Judges agreed with the recommendations in 77 percent of the cases. The judges changed their decisions during the course of reading 35 percent of the 157 modified reports. Figure, tables, and 13 references.