The effects of gain-loss message framing on breast-cancer-related cognitions and behaviors were assessed among 539 women aged 30 to 70 years. The design involved a prebrochure telephone interview, followed by a brochure mailout, and a postbrochure telephone interview. The brochures contained information about breast cancer and the risk of family history. Recommended behaviors were framed to emphasize gains, losses, or were neutral; and statistical risk information was presented either positively or negatively. Measures included demographics, family history, breast self-examination (BSE) performance, BSE intention, self-efficacy in performing BSE, perceived early detection risk of breast cancer, perceived susceptibility to breast cancer, and anxiety about breast cancer. A loss-framed message led to greater positive change in BSE behavior. Interactions between framing effects and variables of issue involvement, perceived early detection risk, and self-efficacy indicated effects on behavior, but not beliefs.