ALL THEORIES OF DETERRENCE PREDICT A NEGATIVE ASSOCIATION BETWEEN AGGREGATE CRIME RATES AND SANCTIONS, WITH LEVELS OF SANCTIONS MEASURED EITHER BY SEVERITY OR BY RISK. EXPERIMENTAL STUDIES OF DETERRENCE ARE RARE; QUASIEXPERIMENTS ARE MORE COMMON. THE MOST COMMON RESEARCH APPROACH IS ANALYSIS OF THE NATURAL VARIATION IN CRIME RATES AND SANCTION LEVELS. SOURCES OF BIAS IN ANALYSES OF NATURAL VARIATION RELATIVE TO NONCAPITAL SANCTIONS INCLUDE ERRORS IN MEASURING CRIME, CONFUSION OF DETERRENCE WITH INCAPACITATION, AND SIMULTANEOUS EFFECTS. ALTHOUGH ANALYSES OF NATURAL VARIATION USUALLY FIND A NEGATIVE ASSOCIATION BETWEEN CRIME RATES AND NONCAPITAL SANCTION RISKS, EXISTING EVIDENCE DOES NOT WARRANT AN AFFIRMATIVE CONCLUSION REGARDING THE DETERRENT EFFECTS OF NONCAPITAL SANCTIONS. THE RESULTS OF EXPERIMENTAL AND QUASI-EXPERIMENTAL STUDIES OF THE DETERRENT EFFECT OF NONCAPITAL SANCTIONS ARE MIXED. IN MOST CASES, THE RESEARCH DESIGNS SUFFER FROM FLAWS THAT UNDERMINE CONFIDENCE IN THE RESULTS. A REVIEW OF EVIDENCE ON CAPITAL SANCTIONS AND DETERRENCE LEADS TO THE CONCLUSION THAT AVAILABLE STUDIES PROVIDE NO USEFUL EVIDENCE ON THE DETERRENT EFFECT OF CAPITAL PUNISHMENT VIS-A-VIS HOMICIDE. EXISTING MODELS FOR ESTIMATING INCAPACITATIVE EFFECTS ON SANCTIONS ARE USEFUL FOR RELATIVE COMPARISONS BUT CANNOT BE RELIED UPON FOR EXACT CALCULATIONS. RECOMMENDATIONS ARE PROVIDED REGARDING RESEARCH IN THE AREAS OF NATURAL VARIATION ANALYSIS, CAPITAL AND NONCAPITAL SANCTIONS, AND INCAPACITATION. RECOMMENDATIONS ARE ALSO PROVIDED REGARDING MANAGEMENT AND ORGANIZATION OF RESEARCH PROGRAMS, WITH ATTENTION TO COMMON DATA NEEDS, ACCESSIBILITY OF DATA, AND PEER REVIEW. (LKM)