This general information volume contains a description of the study background, objectives, operational definitions, and methods. It also includes the results of a literature review and the general conclusions drawn from the study. The first four volumes of the series focus on on major decision categories, including pretrial release, probation and parole level of supervision, institutional custody, and sentencing and parole release. The basic survey design consisted of three distinct elements: literature review, telephone survey, and on-site visits. During the project, the staff made over 350 telephone contacts with experts, agencies, and research organizations throughout the criminal justice system. Twenty-five site visits and numerous structured telephone interviews were completed. The literature review revealed that initial attempts at classification focused on a single decision point (parole release) and were based on a single approach (actuarial prediction of recidivism). Only within the last 15 years have classification instruments been used at other decision points. Although the data do not permit precise assessment, it is clear that the use of classification instruments in criminal justice has increased substantially in recent years. Beginning in the early 1960's, instruments such as the Vera Point Scale were introduced. Moreover, similar kinds of point scales have been more recently adopted for determining eligibility for diversion programs and for making institutional custody-level classifications. The area of parole classification has also seen significant growth since the 1960's. The three main factors involved in the current upsurge of interest in classification instruments are the increasing volume of cases processed by the criminal justice system, the desire to structure discretion in decisionmaking, and the fact that decisions based on instruments are often easier to justify when challenged. Despite over 50 years of reserach aimed at predicting criminal behavior, prediction remains an inexact science. Footnotes, a table, and about 15 165 references are included. The survey questionnaire and site visit protocol are appended. For related documents, see NCJ 76058-61.