THE RELATIONSHIP BETWEEN PRICE PER GRAM OF HEROIN AND THE QUANTITY AND PURITY OF THE PURCHASE AND THE MONTH IN WHICH IT TOOK PLACE IS ESTIMATED ON A CITY-BY-CITY BASIS. THESE ESTIMATED RELATIONSHIPS ARE USED TO INFER THE PRICE PER GRAM WHICH PREVAILED IN THAT CITY FOR EACH MONTH FOR A STANDARDIZED PURCHASE OF TWO GRAMS AT A PURITY LEVEL OF 10 PERCENT. LABORATORY ANALYSIS DATA PROVIDED ENOUGH INFORMATION FOR MONTHLY HEROIN PRICE INDICES TO BE CONSTRUCTED FOR SOME PART OF THE TWO-YEAR PERIOD FROM JULY 1970 TO JUNE 1972 FOR NEW YORK CITY, BOSTON, BALTIMORE, MIAMI, NEW ORLEANS, HOUSTON, CHICAGO, DETROIT, AND LOS ANGELES. RETAIL PRICE ESTIMATES, BASED UPON PURCHASES OF HEROIN BY NARCOTICS ENFORCEMENT AGENTS, SHOWED CONSIDERABLE FLUCTUATION OVER TIME, A FACT THAT HAD NOT BEEN REPORTED SYSTEMATICALLY BEFORE. APPLICATIONS OF THE ESTIMATED PRICE SERIES PROVIDED INSIGHT INTO THE STRUCTURE OF THE HEROIN MARKET AND THE RELATIONSHIP BETWEEN HEROIN PRICES AND CRIME. IN NEW YORK CITY, CRIME IN REVENUE-RAISING CATEGORIES APPEARS TO FLUCTUATE SIGNIFICANTLY WITH THE RETAIL PRICE OF HEROIN. EXTENSIONS OF THE ANALYSIS TO OTHER POLICY-RELATED QUESTIONS, SUCH AS SUBSTITUTABILITY AMONG DRUGS, ENTRY INTO TREATMENT PROGRAMS, AND EFFECTIVENESS OF LAW ENFORCEMENT POLICIES, ARE DISCUSSED. APPENDED MATERIALS INCLUDE SUPPLEMENTAL DATA FROM THE LABORATORY ANALYSIS REPORTS, A DISCUSSION OF MATHEMATICAL MODELS FOR ESTIMATING THE STANDARDIZED 'RETAIL' PRICE SERIES FOR HEROIN, AND FURTHER RESULTS OF THE ANALYSIS OF THE RELATIONSHIP BETWEEN DRUG PRICES AND CRIME. A REFERENCE LIST IS PROVIDED. (AUTHOR ABSTRACT MODIFIED)