The study aimed to develop a severity scale using magnitude estimation techniques, examine the consensus among different populations, test the additivity and linearity of such a scale, examine the severity of the death penalty in comparison with other penalties, and test a variety of methodological issues. The subjects were 15 police officers, 15 inmates at the Philadelphia House of Corrections, 26 probation officers, and 28 undergraduates in a criminology class at the University of Pennsylvania. The subjects considered lists of 36 penalties spread over 3 pages in random order. Two additional questionnaires were used to test the significance of the numbers used in scoring and the randomization of the order of the items. Probation and fines were considered less severe sanctions than imprisonment, except where heavy doses of these measures were used. Highest values were recorded by the prisoners, the next highest by the students, the third highest by the probation officers, and the lowest by the police officers. Some validation was found for the Hamilton and Rotkin hypothesis as to the special nature of the death penalty. Additional findings and policy implications are included, together with figures, data tables, 31 footnotes, and 29 references. An appendix presents the study instrument.