A case-control study was undertaken to determine the dietary intake of vitamin A in women having abnormal uterocervical cytology. The study groups (87 cases and 82 controls) were drawn from a population of women who received a screening Pap test in the ambulatory health care section of a large municipal hospital center. A subset of cases (with abnormal cytology) were matched to controls for age, ethnicity, socioeconomic status, and parity. Nutrient intake and retinol binding protein concentrations were determined; epidemiological data were also obtained. It was found that the subset of cases with severe dysplasia or carcinoma in situ (CIS) were more likely to have a total dietary vitamin A intake below the pooled median (3,450 IU) and/or a beta-carotene intake below the pooled median (2,072 IU) than were normal controls (p less than 0.05 and p less than 0.025, respectively). Odds ratios revealed approximately a 3-fold greater risk for severe dysplasia or CIS in women with lowered vitamin A or beta-carotene intake. In addition, retinol binding protein was either absent or undetectable in 78.8% of the dysplastic tissue samples, versus 23.5% of the normal tissue samples (p less than 0.005).