The survey measures the public's assessment of the severity of 204 crime vignettes, including traditional offenses covered in existing crime statistics as well as a number of nontraditional crimes. The study uses a magnitude method of estimation, in which a respondent can assign any value to any item -- the scale has no upper limit. Geometric means and ratio scores are calculated from the values assigned by respondents. Chapters 1 through 4 of this report describe the study background and include detailed discussion of a number of methodological issues that were considered in developing the scale, including validation of the scale, the problem of additivity, and relationship between perceived severity of an offense and offender-victim weapon types. Chapters 5 and 6 describe the pretesting and administration of the survey, including revisions made in the study plan as a result of pretesting, the sample design, and the procedures followed by the U.S. Bureau of the Census in conducting the interviewing. Chapters 7 and 8 present findings from the survey, on both national and regional levels. The effect of demographic variables on assessing crime severity is also discussed. Chapter 9 indicates how the scale can be applied; this chapter includes a number of criminal justice applications. A summary of findings from this study was published in the Bureau of Justice Statistics Bulletin, 'The Severity of Crime,' NCJ-92326, January 1984. Another volume, the 'Sourcebook of Crime Severity Ratios for Core-item Offenses,' 850 pp., NCJ-96239, available only in microfiche because of its length and technical nature, presents tables of crime severity ratios for various demographic characteristics by census regions and divisions, Office of Management and Budget regions, States, Standard Metropolitan Statistical Areas, major cities, and the total U.S. population. (Author abstract modified)