GIVEN THE NEED TO CONTROL ALL THE VARIABLES PRESENT IN DETERRENCE RESEARCH, INVESTIGATORS SHOULD USE AN EXPERIMENTAL RATHER THAN OBSERVATIONAL METHODOLOGY. EVEN THOUGH IT APPEARS DESIRABLE TO REDUCE THE NUMBER OF VARIABLES PRESENT IN SUCH STUDIES, DETERRENCE ONLY RESULTS FROM THE CUMULATIVE AFFECT OF ALL THESE VARIABLES. DETERRENCE RESEARCHERS SHOULD ABANDON CONVENTIONAL CRIME RATES AND USE SPECIAL RATES, SUCH AS THE REPETITIVE RATE. A MAJOR ISSUE CONCERNS THE KINDS OF POPULATIONS APPROPRIATE FOR DETERRANCE RESEARCH SINCE MOST PEOPLE RARELY CONTEMPLATE COMMITTING CRIMES LET ALONE FEEL DETERRED FROM DOING SO. ADDITIONAL COMPLICATIONS ARISE IN CONDUCTING DETERRENCE RESEARCH WHEN CERTAIN NONDETERRENT PREVENTIVE MECHANISMS, SUCH AS IMPRISONMENT, MAY BE CONFUSED WITH DETERRENCE. IN ADDITION, EXTRALEGAL CONDITIONS WHICH AFFECT CRIME RATES, SUCH AS UNEMPLOYMENT AND SOCIETAL ATTITUDES TOWARD CRIME, MUST BE CONSIDERED. MOST STUDIES PRIOR TO 1968, ON THE GENERAL DETERRENCE AFFECT OF PUNISHMENT ON SOCIETY AS A WHOLE, CONCENTRATED ON COMPARISONS OF THE CRIMINAL HOMICIDE RATES FOR STATES HAVING CAPITAL PUNISHMENT AND THOSE WITHOUT CAPITAL PUNISHMENT. WHILE SUCH STUDIES CONTAIN SERIOUS FLAWS, NO EVIDENCE HAS BEEN FOUND TO INDICATE THAT SEVERE PUNISHMENT RESULTS IN A DECLINE IN THE CRIME RATE. ALTHOUGH AN EXPERIMENTAL STUDY OF THE DETERRENCE DOCTRINE WOULD BE COMPLICATED AND COSTLY, THERE IS NO COMPELLING REASON WHY IT CANNOT BE USED. HOWEVER, MOST EXPERIMENTAL RESEARCHERS ARE UNINTERESTED IN DETERRENCE DOCTRINE RESEARCH AND ARE UNSKILLED IN EXPERIMENTAL METHODOLOY. A DIAGRAM, NOTES, AND REFERENCES ARE PROVIDED. (KCP)