Data from a 1% sample of households in the 1980 United States Census of Population and Housing were used to identify personal and social characteristics associated with being poor or marginally poor in Colorado. The general hypothesis examined was that the more limited an individual's access is to participation in the nontraditional aspects of the economic and social opportunity structure of the state and society, the more likely he/she is to be impoverished. Disproportional poverty is thus expected in remote, traditional areas and among people with traditional social identities that limit access to economic opportunities. The following variables were considered: area of the state, age, gender, minority status, English language skill, education completed, present school enrollment, disability status, marital status, responsibility for dependent children, rural and farm residence, recent migration history, whether employed, occupation type, and industry category according to traditionality or recency in Colorado. With some qualifications, the data supported the research hypothesis. Of the variables examined, location, marital, minority, and employment status made the greatest difference statewide. Education level was important in traditional rural areas, and current school enrollment was important in Colorado suburban, small SMSA (Standard Metropolitan Statistical Area), and nontraditional rural areas. Implications for programs and policies are discussed. (JHZ)