This case-control study addressed the hypothesis that uninterrupted exposure to light is associated with increased rates of breast cancer. We compared the odds of profound binocular blindness among women with a diagnosis of breast cancer with the odds of profound binocular blindness among women with diagnoses of coronary heart disease or stroke. All hospital discharges in the National Hospital Discharge Survey from 1979 through 1987 were analyzed, after exclusion of women with diabetes. Profoundly blind women were half as likely to have breast cancer as women who were not profoundly blind. This effect diminished substantially with increasing age.