A THEORETICAL MODEL IS PRESENTED WHICH EMPLOYS THE RICARDIAN ASSUMPTION OF AN INELASTIC SUPPLY OF LAND IN CONJUNCTION WITH THE EFFECT OF CRIME ON THE PRICE OF LAND. INSOFAR AS CRIME COSTS CAN BE MEASURED BY SOME IMPLICIT MARKET, THEY ARE LIKELY TO BE CAPTURED BY LAND AND NOT CAPITAL PRICES. THE MODEL IS APPLIED TO DATA FROM CHICAGO'S HYDE PARK-KENWOOD COMMUNITIES FOR A BLOCK-BY-BLOCK ANALYSIS. THREE TYPES OF DATA ARE USED AS DEPENDENT VARIABLES--1970 CONTRACT RENTS OF RENTER-OCCUPIED HOUSING (FROM THE CENSUS OF HOUSING) FOR 11 BLOCKS; 324 PRICES (85 BLOCKS) OF SINGLE-FAMILY HOMES TRANSACTED OVER THE YEARS 1969 TO 1972; AND OWNER-ESTIMATED 1969 VALUES OF OWNER-OCCUPIED HOUSING FOR A TOTAL OF 68 BLOCKS. EXTENSIVE CRIME DATA WERE COLLECTED AND USED TO COMPUTE A RATE OF CRIME-ON-THE-BLOCK AND A RATE FOR THE AVERAGE OF ALL THE CONTINGUOUS BLOCKS. THE CRIME CATEGORIES INCLUDED TOTAL INDEX CRIME, CRIMES AGAINST PERSONS, AND CRIMES AGAINST PROPERTY. A TRANSIENT POPULATION DUMMY VARIABLE WAS ENTERED INTO REGRESSION EQUATIONS TO APPROXIMATE MORE CLOSELY CHANGES IN TRUE CRIME RATE. THE FOLLOWING NONCRIME VARIABLES WERE ENTERED INTO ESTIMATING EQUATIONS--LOG OF NUMBER OF RENTER ROOMS, PROPORTION OF HOUSING BUILT AFTER 1960, MEDIUM FAMILY INCOME, DISTANCE FROM THE UNIVERSITY, PROPORTION OF WHITES, EAST HYDE PARK REGION, PARK REGION, AND TRANSIENT POPULATION. THE CRIME COST IS THEN COMPUTED FOR BOTH THE HYDE PARK AREA AND THE WHOLE OF CHICAGO. RESULTS REVEAL THAT THE CRIME RATE IS HIGHER IN THE HYDE PARK-KENWOOD COMMUNITY THAN IN CHICAGO AS A WHOLE AND THAT THE COST PER CRIME IN THIS DISTRICT IS 48 PERCENT HIGHER. HOUSING VALUE, RACIAL CONCENTRATION IN NEIGHBORHOODS, AND NUMBER OF RENTER AND OWNER-OCCUPIED UNITS ARE SIGNIFICANT FACTORS. TABULAR DATA, THE STUDY LOGARITHMS, AND REFERENCES ARE PROVIDED. (DAG)