THE MATHEMATICAL MODEL USED TO RELATE THE LEVEL OF CRIME TO THE PRICE OF HEROIN WAS AS FOLLOWS: CRIME LEVEL=F (HEROIN PRICE, HEROIN POTENCY, OFFENSE CLEARANCE RATE, TEMPERATURE, SEASON, TIME TREND). ALL VARIABLES WERE MEASURED MONTHLY AND RELATED OVER THE TIME PERIOD FROM NOVEMBER 1970 TO JULY 1973 USING TIME SERIES REGRESSION ANALYSIS. LAGGED VALUES OF HEROIN PRICE AND OFFENSE CLEARANCE RATE WERE USED ALONG WITH THEIR CURRENT MONTH VALUES. ANALYSIS WAS CARRIED OUT AT BOTH THE CITY-WIDE AND NEIGHBORHOOD LEVELS. STUDY RESULTS SHOWED THAT PROPERTY (I.E., REVENUE-RAISING) CRIME INCREASES WHEN THE PRICE OF HEROIN RISES, BUT THAT THE SIZE OF THE INCREASE VARIES ACROSS THE CITY. IT WAS ALSO FOUND THAT BY INCREASING THE EFFECTIVENESS OF THE POLICE IN ARRESTING OFFENDERS, THE LEVEL OF CRIME CAN BE REDUCED. APPENDED MATERIALS INCLUDE A DETAILED EXPLANATION OF THE MATHEMATICAL MODEL, TABULAR DATA ON THE PRICE ELASTICITIES OF PROPERTY CRIME FOR EACH OF 41 NEIGHBORHOODS, EMPIRICAL FINDINGS ON THE ADJUSTMENT OF CONSUMPTION TO PRICE CHANGES, AND MISCELLANEOUS CRIME FIGURES BY NEIGHBORHOOD. FOR A RELATED PAPER, SEE NCJ-37314.