Increased exposure to endogenous estrogen and/or insulin may partly explain the relationship of obesity, physical inactivity, and alcohol consumption and postmenopausal breast cancer. However, these potential mediating effects have not been formally quantified in a survival analysis setting.