THE PRIMARY CONCERN OF THE ANALYSIS IS TO DETERMINE WHETHER CAPITAL PUNISHMENT PROVIDES A DETERRENT EFFECT INDEPENDENT OF ANY DETERRENT EFFECT OF THE THREAT OF IMPRISONMENT. THE CRIMES EXAMINED INCLUDE HOMICIDE, RAPE, AND ASSAULT. PROBABILITY OF IMPRISONMENT IS EXPRESSED AS THE RATIO OF PRISON COMMITMENTS TO REPORTED OFFENSES OF A GIVEN TYPE. THE LENGTH OF PRISON TERM VARIABLE IS A MEASURE OF THE AVERAGE TIME SERVED BY OFFENDERS FOR EACH CRIME TYPE. PROBABILITY OF EXECUTION IS REPRESENTED IN SEVERAL DIFFERENT FORMS. THE MODEL INCLUDES CONTROLS FOR PER CAPITA POLICE EXPENDITURES, PERCENTAGE OF BLACKS IN THE POPULATION, AGE, REGION, UNEMPLOYMENT, AND MEDIAN INCOME. THE FINDINGS SUPPORTING THOSE REPORTED BY OTHER RESEARCHERS USING CROSS-SECTION DATA SETS. THERE APPEAR TO BE NO INDEPENDENT DETERRENT EFFECTS OF CAPITAL PUNISHMENT FOR VIOLENT CRIME WHEN PROBABILITY OF IMPRISONMENT, LENGTH OF PRISON TERM, AND OTHER CONTROL VARIABLES ARE INCLUDED IN THE MODEL. THE FINDINGS DO SUGGEST THAT THE THREAT OF IMPRISONMENT DETERS BOTH HOMICIDE AND RAPE. ALTHOUGH ASSAULT APPEARS NOT TO BE RESPONSIVE TO PUNISHMENT VARIABLES, THE FINDINGS FOR HOMICIDE INDICATE THAT THE MORE SEVERE HOMICIDE PENALITIES (OTHER THAN CAPITAL PUNISHMENT) MAY SERVE TO PREVENT ASSAULTS FROM ESCALATING INTO HOMICIDES. TABLES AND A LIST OF REFERENCES ARE INCLUDED. (LKM)