After the Supreme Court decision in Furman v. Georgia (1972), many States enacted new death penalty legislation. These statutes applied the death penalty to murder of an on-duty correctional employee, murder by a life-term prisoner, murder by any prisoner, murder by a convicted capital offender, and/or murder by a convicted murderer. Retribution and deterrence were cited as reasons for the new legislation. The rationale based upon the deterrence argument can be further clarified by empirical testing. One comprehensive study of prison homicides in State and Federal prisons was conducted in 1973, a year when 124 persons were murdered in prison. Many of the findings raise questions about assumptions traditionally associated with the deterrence argument. According to study findings, the inmate's risk of being killed in prison did not exceed the victimization rate for criminal homicide in the general population outside prison. Likewise, for prison staff the hazards of prison life have probably been overestimated. In general, findings argue that the threat of the death penalty fails to deter behavior prohibited outside prison just as it failed in 1973 to deter 94 percent of the prison homicides. Deterrence tactics should be abandoned in favor of measures designed to reduce the risk of prison murder. Lethal weapons should be removed from the prison inmate population. Cells should be occupied by only one prisoner, and defects in the structural design of prisons should be corrected. Prison staffs should be educated to recognize a potentially fatal injury. The article provides 5 tables and 43 footnotes.