Ambulatory care sensitive admission rates have been proposed as measures of access to health care. To test this, admissions for ambulatory care sensitive conditions (ACSC) were analyzed by multiple linear regression. The percentage of population below 200 percent of the federally defined poverty level, the percentage of black people, and the number of primary care providers per 1,000 population were found to be positively associated with ACSC admissions. Population density was negatively associated with ACSC admissions. There was no association between the location of the ZIP code in a health professional shortage area and ACSC admissions. Proximity to the hospital was found to be positively associated with ACSC admissions but was examined only in the most rural ZIP code group. The significant independent variables and the direction of their effects were the same across all ZIP code groups. The analysis suggests that high ACSC admissions may be a reflection of deficits in one or more of the following areas: primary care availability, accessibility, or appropriateness. In-depth study is needed to determine the relative importance of these factors in a given geographical area. There also may be environmental and social factors external to the health care system that contribute to ACSC admissions. The findings suggest that ACSC should be used cautiously as a measure of primary care system needs, and in conjunction with other health, demographic, or service utilization data.