The NCS data, collected by the Census Bureau for 1975 and 1976, were used in the study. A sample of 54,343 homes, of which 2,133 were victimized by residential burglary, were studied. A model-fitting and selection procedure were used to estimate effect parameters. The variables used were race, age, income, area type, household occupancy, and whether or not those sampled reported a burglary during the period under study. The study showed that households headed by the central city residents, the young, persons with incomes higher or lower than average, nonwhites, and persons whose homes are unoccupied relatively often are most likely to be burglarized. The types of persons heading households least likely to be burglarized were older citizens, persons residing outside the central city, citizens in middle income categories, and persons whose homes were occupied relatively often. These profiles could be explained by a theory of two types of burglars. One type could reside in the low-income areas of central cities and outlying areas and capitalize on situational opportunities to burglarize the very poor. Conversely, professional burglars would prey upon the very rich in both the central city and other areas. These findings confirmed the importance of occupancy in the selection of burglary targets. The relationship between the age of the head of the household and the risk of burglary victimization provided additional support for the importance of this factor, since older persons are home more often. These data also indicated that the key to a reduction in burglary rates might be tighter social organization of neighborhoods. Prior studies are discussed; statistical data, footnotes, and 19 references are included.