As a means of developing effective intervention strategies for promoting Pap smear screening, we analyzed data from a population-based women's health survey (N = 603) in a 36-county area in southeastern Kentucky. The cervical cancer mortality rate for white women in this area is one of the highest in the United States. By using selected sociodemographic, health-care utilization, health knowledge, and health behavior variables in age-specific logistic regression models, we discriminated between women who had had a Pap smear within 3.5 years and those who had not. Several variables predicted Pap screening status regardless of the woman's age. Women of all age groups who had not been recently screened had encounters with the medical-care system. A key variable that affected use of screening services was ever use of birth control pills. The main differences between the three age groups were as follows: the 18-44 age group was less likely to see a private physician and less likely to seek medical care of any type, except for care related to pregnancy; only the 45-59 age group believed that cost of medical care was a problem; and only for the 60 or older age group were socioeconomic variables associated with not having recently had a Pap test.