The first known study of victimization dates back to 1720 in Denmark, but victim surveys in their present form can be traced to the 1960's. Innovations such as inclusion of the dark figure, comprehensive geographical coverage, and special surveys of large cities, were introduced by the U.S. Law Enforcement Assistance Administration and the Bureaus of the Census. There are two main types of local victim surveys. Comprehensive studies measure the extent of reported and unreported index offenses in a particular geographical unit, and gauge public attitudes and trends in buying habits, moves, and fear of crime. Limited studies collect data on specific offenses (e.g., wife beating). Limited studies are intended to facilitate planning of assistance programs and to evaluate the effectiveness of preventive efforts. Problems such as validity checks and distortions from time lags between crimes and interviews and from telescoping of time in memory are solved with the reverse police reports check and bounded interviews. Other important problems are establishment of a period of deference, use of jargon and long questions, and the need to interview several family members for accurate information. The recent technique of random digit dialing, i.e., use of randomly selected telephone numbers for collecting interview information has proved less costly than personal interviews. Difficulties such as establishing statistics for victimless crimes and interviewing corporate victims must yet be resolved. Victim surveys intend to measure police effectiveness, crime rates, and usefulness of preventive programs, to determine victimization risks, to ascertain public attitudes toward crime and police, and to facilitate planning. Results of significant local studies since 1966 have shown that crime rates are twice as high as expected from police records, that two-thirds of all offenses involve offenders and victims who are strangers, that minor property crimes are most common, and that theft is the most usual offense against the elderly. Studies of child abuse, rape victims, and violence in marriage have relied heavily on police records because of difficulties in interviewing victims and the sensitivity of the problems involved. Notes and a 48-item bibliogrphy are supplied.