An introduction concludes that subjective deliberations, shaped by subcultural norms, result in decisionmakers judging carelessly or in a biased fashion. Although statistical tabulations that offer ways of identifying relative risk from past experience were extolled as a remedy to this problem, initially they were little used in criminal justice agencies. An accelerating rate of acceptance of statistical tables for risk classification occurred during the 1970s and 1980s because of increasing collaboration between researchers and criminal justice officials. The following applications of risk classification guidelines are described: pretrial release, selecting cases for prosecution, sentencing guidelines, custodial risk classification, parole decisions, levels of probation and parole supervision, and quasi-experimental evaluation. A concluding discussion emphasizes that decision guidelines are not used nearly as must nor as well as they should be. Tables, charts, and approximately 90 references. (Author abstract modified)