INITIATED IN 1972, THE NCS IS A CONTINUOUS REPRESENTATIVE SURVEY OF HOUSEHOLDS, CONDUCTED, TABULATED, AND ANALYZED BY THE U.S. BUREAU OF THE CENSUS FOR LEAAA. IT PROVIDES ANNUAL ESTIMATES ON THE EXTENT OF CRIME AGAINST INDIVIDUALS AGED 12 AND OVER AND FOR HOUSEHOLDS. IN ADDITION, THE SURVEY FURNISHES VICTIM CHARACTERISTICS AND DETAILED INFORMATION ON THE CIRCUMSTANCES OF VICTIMIZATION. ALTHOUGH THE NCS PROVIDES A LARGE SAMPLE AND RESULTING SMALL STANDARD ERRORS FOR HYPOTHESIS TESTING, SPECIFIC INVESTIGATIONS OF SUCH ISSUES AS DOMESTIC VIOLENCE, JUVENILE VICTIMS, OR REPORTING OF CRIME TO THE POLICE BY MINORITY GROUPS BECOME PROBLEMATICAL BECAUSE OF THESE (SAME) SAMPLE SIZE AND STANDARD ERROR CONSIDERATIONS. AN ADDITIONAL PROBLEM DERIVES FROM THE NCS FINDING OF NONDISCRETE EVENTS OR SERIES CRIMES WHERE VICTIMIZATIONS ARE SIMILAR IN DETAIL, CONSTITUTE AT LEAST THREE IN A SERIES, AND THE NOT REPORTED SEPARATELY BY THE RESPONDENT. THE EXCLUSION OF SERIES CRIMES FROM THE NCS CRIME ESTIMATES THEREFORE, PRESENTS PROBLEMS OF REDUCED ESTIMATED CRIME RATES, AFFECTS RELATIVE DISTRIBUTIONS OR RATIOS OF OFFENSES BY TYPE, AND PROBABLY DISTORTS VICTIM AND CRIME-EVENT RELATIONSHIPS FOR NONSERIES DATA. NCS PROVIDES THREE TYPES OF AREA DATA: (1) STATE-LEVEL STATISTICS COMPARABLE TO THE ANNUAL NATIONAL STATISTICS FOR 1974 THROUGH 1976 IN THE 10 LARGEST STATES; (2) DATA FROM STANDARD METROPOLITAN STATISTICAL AREAS (SMSA'S) THAT FOCUS ON THE LOCALITY OF THE VICTIM AND COMPARE THE AGGREGATE PATTERNS OF CRIME AGAINST CENTRAL CITY, SUBURBAN, AND NONMETROPOLITAN RESIDENTS; AND (3) DATA BASED ON NEIGHBORHOOD CHARACTERISTICS. ANALYTIC PROBLEMS INCLUDE MATCHING NCS NEIGHBORHOOD CHARACTERISTICS WITH THE LOCALITY OF HOUSEHOLDS, NOT ACCIDENTS; SELECTION OF AT-HOME INCIDENTS THAT COULD BIAS ANALYSIS; AND BIAS OR ERROR INTRODUCED BY REPLACEMENT HOUSEHOLDS IN HOUSEHOLD BURGLARY AND LARCENY STATISTICS.