BENEFIT-COST ANALYSIS HAS GENERAL LIMITATIONS, SUCH HAS ITS INABILITY TO DEAL WITH EQUITY CONSIDERATIONS. ONE OF ITS MAIN LIMITATIONS IN CRIMINAL JUSTICE RESEARCH IS THE DIFFICULTY OF OBTAINING DEFINITIVE ESTIMATES OF MANY CRIMINAL JUSTICE BENEFITS AND COSTS. IN THE PRESENT ILLUSTRATION, TWO ALTERNATIVE SENTENCES (EXECUTION AND LIFE IMPRISONMENT) WERE CONSIDERED. COSTS INCLUDED LOSS IN OUTPUT DUE TO INCAPACITATING THE CRIMINAL AND COSTS OF IMPLEMENTING THE SENTENCE. EXECUTION HAD THE ADDITIONAL COSTS OF LIVES TAKEN INCORRECTLY AND OF THE TAKING OF LIFE BY THE GOVERNMENT. BENEFITS INCLUDED RETRIBUTION, DETERRENCE, PUNISHMENTS, SAVINGS IN OTHER SOCIAL PROGRAMS, OTHER CRIMES AVERTED, AND CRIMINAL JUSTICE SYSTEM COSTS AVOIDED. UNLIKE MOST ANALYSIS, THE PRESENT STUDY TREATED THE DISCOUNT RATE AS A RANDOM VARIABLE WHICH WAS DISTRIBUTED NORMALLY. USING ESTIMATES OF VALUES OF EACH BENEFIT AND COST, 15 BENEFIT-COST RATIOS WERE COMPUTED FOR EACH SENTENCING ALTERNATIVE AND DIFFERENCES BETWEEN THE ALTERNATIVES WERE COMPARED. RESULTS SHOWED THAT SINGLE POINT ESTIMATES YIELDED DIFFERENT RESULTS FROM THE GROUPS OF ESTIMATES. ANALYSIS OF ALL ESTIMATES SHOWED EXECUTION TO BE SUPERIOR TO LIFE IMPRISONMENT FROM THE STANDPOINT OF ECONOMIC EFFICIENCY. DIFFERENCES BETWEEN THE GROUP MEANS WERE SIGNIFICANT AT THE .005 LEVEL. BECAUSE OF THE NUMBER OF ARBITRARY DECISIONS, THE ANALYSIS DOES NOT CONCLUDE, EVEN ON EFFICIENCY GROUNDS, THAT EXECUTION IS SUPERIOR TO LIFE IMPRISONMENT BUT, RATHER, THAT USING STATISTICAL METHODS IN BENEFIT-COST ANALYSIS (WHICH IS SUBJECT TO MARKED UNCERTAINTY IN VALUATION OF DATA) CAN GREATLY INCREASE THE TECHNIQUE'S USEFULNESS. TABLES, NOTES, AND A REFERENCE LIST ARE INCLUDED. (CFW)