Data collected by the Victoria Police at the time an alleged drink-driving offender undergoes an evidential breath test for the presence of alcohol were analyzed to investigate the possibility that occupational information could be used to define groups of drink-drivers with similar characteristics. Such groups could then be utilized in the development of targeted public education campaigns in the Victorian (Australia) context where there are already high levels of enforcement and mass-media publicity. It is argued that drink-drivers in this context are likely to share some characteristics which might reduce the impact of current programs. Analysis of the data relied on the application of a theory of the relationship between career choice and other behavioral characteristics (Holland, 1973, Making Vocational Choices, Prentice Hall, Englewood Cliffs, N.J.; Holland, 1975, Manual for the Vocational Preference Inventory, Consulting Psychologists Press, Palo Alto, CA). The analysis identified two potential groups which accounted for 58% of the male drink-drivers. These groups differed from other male drink drivers in a number of ways, underlining the potential for their use as targets in future campaigns. One occupational category accounted for 42% of the male drink drivers, including occupations such as carpenter, electrician, chef, mechanic, gardener, and laborer. The behavioral characteristics associated with these occupational codes in the Holland model included asocial, conforming, reserved, introspective, unpopular, orderly, careful, unimaginative, and defensive. The other occupational category accounted for 16% of the male drink drivers and included occupations such as business manager, company director, public servant, and sales representative. Behavioral characteristics associated with this combination included acquisitive, adventurous, ambitious, energetic, extroverted, friendly, and generous.