We conducted a population survey to describe patterns and determine predictors of the use of nutritional supplements and single-ingredient vitamins and minerals among elderly living in five adjacent urban and rural counties in the Piedmont area of North Carolina. The stratified random sample consisted of 3939 black and white participants age 65 or older from the Duke Established Populations for Epidemiologic Studies of the Elderly. The use of nutritional supplements within the previous 2 weeks was determined during an in-home interview. Multivariate analyses, using weighted data adjusted for sampling design, were conducted to assess the association between nutritional supplement use and predisposing, need, and enabling factors. Nutritional supplement use was reported by 26.2% of participants and was more likely for those who were white women, were high school educated, were underweight, took prescription drugs, had five or more health visits in the previous year, and had supplemental health insurance. It was less likely for those with poor self-rated health. The majority (71.5%) of nutritional supplement users took at least one single-ingredient supplement. Use of such products was more likely in those who were white, born and raised in an urban area, and high school educated, and was less likely in those with impaired functional status. Nutritional supplement use is prevalent in community-dwelling elderly and is more commonly associated with demographic factors and access to health care than with need factors.