The relationship between an index of dietary vitamin A and cancer risk at 25 sites was assessed in retrospective case-control studies. Common control groups for males and females were used in all analyses. Relative risk estimates were derived from multiple logistic regression analyses that controlled for age, alcohol consumption, and smoking exposure. We found that among males, dietary vitamin A is associated with lower risk for cancers of the tongue, floor and other mouth, pharynx, larynx, esophagus, and lung but higher risk for Hodgkin's disease and leukemia. Among females, we found that dietary vitamin A had less effect on risk generally but was associated with lower risk for bladder cancer. These findings are consistent with previous research that showed dietary vitamin A to be associated with decreased risk of squamous epithelial cancers. The association of dietary vitamin A and increased risk of Hodgkin's disease and leukemia among males in addition to the disparity in effect of dietary vitamin A on risk between males and females are areas worthy of further research.