The study was designed to test the basic thesis of the routine activity approach which suggests that variations in individual daily activity increase or lessen opportunities for crime. Log-linear models of contingency table analysis were used along with data collected through the National Crime Survey (NCS) and by the Census Bureau. These data offered several advantages over data used in previous studies: the NCS samples the entire resident, noninstitutionalized population of the United States, and the sample of over 100,000 citizens permits a comparison of victims and nonvictims and a simultaneous control over more variables than would be possible in a smaller survey. The effects of age, race, income, major activity, and number of household persons were examined independently and in interaction. Data indicated that those with a family income of less than $10,000 a year, those who were 50 years or older, and those whose major occupation was 'keeping house,' had less than average odds of being the victims of a personal larceny. Furthermore, persons with a family income of more than $20,000, who live alone, or are between the ages 16 and 29 face a greater than average risk of being larceny victims. These findings indicated substantial support for the routine activity perspective and its promise as a base from which to develop a general theory of criminal victimization. Footnotes, 18 references, and tabular data are provided. (Author abstract modified).