Obesity, alcohol consumption, physical inactivity and postmenopausal hormone use are known modifiable risk factors for breast cancer. We aim to measure incidence rates of breast cancer for women with favorable levels on all 4 risk factors (BMI <or= 30 kg/m(2), alcohol <1 drink/week, physically active and no current hormone use) and to evaluate their associations with estrogen. The 5,054 postmenopausal women in the Copenhagen City Heart Study were asked about risk factors at baseline in 1981-3 and were followed until 2002 in the Danish Cancer Registry, with <0.1% loss to follow-up. Estradiol was measured in a subset of 1,042 women. During follow-up, 263 women developed breast cancer. Twenty-six percent of the women had a favourable risk factor profile, and their breast cancer rates were markedly lower (154 per 100,000 years) than women with 3+ risk factors (460 per 100,000 years). One, two and three risk factors were associated with hazard ratios of 1.38 (95% CI: 0.99; 1.92), 1.84 (1.26; 2.67) and 2.79 (1.59; 4.88) compared to women with a favourable profile. Each of the risk factors was associated with estrogen. In conclusion, the risk of breast cancer was markedly lower for women with a favourable risk profile than for other women and lower estrogen levels is a possible explanation.