ANALYSIS OF REPORTED RESEARCH ON CRIME DETERRENTS INDICATES THAT THE CERTAINTY OF PUNISHMENT SEEMS TO BE A MORE EFFECTIVE DETERRENT THAN THE NATURE OF SEVERITY OF THE PUNISHMENT. NO UNIVERSAL DETERRENT EXISTS, AND THE SUCCESS OF ANY PARTICULAR PUNISHMENT LARGELY DEPENDS ON THE TYPE OF BEHAVIOR TO WHICH IT IS APPLIED. THE CLASS OF PERSONS TO BE DETERRED, AND THE CIRCUMSTANCES IN WHICH THE PERSONS ACT. BECAUSE LONGER SENTENCES ARE INEFFECTIVE AS DETERRENTS, THE SCALING DOWN OF MAXIMUM SENTENCES AND OF THE AVERAGE PRISON SENTENCE COULD BE ACCOMPLISHED WITHOUT LEADING TO HIGHER CRIME RATES. THE FIRST FEW WEEKS OF A PRISON SENTENCE CONTRIBUTE TO DETERRENCE, AFTER WHICH A HARDENING OF ATTITUDES AND AN INCREASING RESENTMENT OCCUR. CAPITAL PUNISHMENT IS NOT JUSTIFIED ON DETERRENT GROUNDS BECAUSE THE ALTERNATIVE OF A LONG PRISON SENTENCE IS NO LESS A THREAT. IN ADDITION THE INTRODUCTION OF MORE SEVERE SENTENCES MAY REDUCE THE LIKELIHOOD OF THEIR BEING IMPOSED AND MAY HAVE A COUNTERPRODUCTIVE NONDETERRING EFFECT. IDEALLY, DETERRENT MEASURES WOULD WORK BEST IF THEY COULD BECOME INCORPORATED WITH STANDARDS OF MORALITY IN THE COMMUNITY, WHICH NEITHER DISCRIMINATED AGAINST INDIVIDUALS NOR CREATED MARTYRS OF THE CONVICTS. THE MOST EFFECTIVE DETERRENTS ARE THOSE WHICH MAKE OBVIOUS THE LIKELIHOOD OF FAILURE AT THE CRIMINAL ACTIVITY. AREAS FOR FUTHER STUDY, INCLUDING THE EFFECTS OF PUBLIC AWARENESS OF PENALTIES, ARE PRESENTED WITH REFERENCES. (TWK)