Abstract: Although chiropractic is used by approximately 10 percent of the U.S. population, predictors of its use have not been definitively described. Previous studies have suggested that chiropractic users differ from nonusers in a number of socio demographic characteristics, but their findings are inconsistent, perhaps because of differences in populations sampled and dates of data collection, most of which are prior to 1990. Regional studies have been conducted in rural areas based on the premise that rural residents are more likely than non-rural residents to use chiropractic; however, this premise has not been definitively documented. The purpose of this study was to provide clarification of these sociodemographic predictors of chiropractic use in Illinois, Iowa, Minnesota, Missouri, Nebraska, South Dakota and Wisconsin by characterizing chiropractic users and nonusers in terms of sociodemographics, including rural or non-rural residence, and presence of low back pain. Data from 1,511 respondents to a 1994 population-based survey, conducted by the University of Iowa Social Science Institute, were analyzed. Unconditional logistic regression was used to derive odds ratios and 95 percent confidence intervals for univariate and multiple regression models. Overall, 15.1 percent of respondents had used chiropractic within the last year, most often for low back pain (57 percent). Chiropractic use was less likely in African Americans, Hispanics and Asians than whites, less likely by non-rural than rural residents, and less likely in Catholics than Protestants in states other than Iowa and South Dakota. Overall, 42.7 percent of workers with low back pain reported using chiropractic, and use increased with age but remained significantly related to race, rural or non-rural residence, state of residence and religious preference. Race, rural or non-rural residence, state of residence and religious preference, independently of low back pain, affect use of chiropractic in seven Midwestern states.