The level of total adiponectin, a mixture of different adiponectin forms, has been reported associated with breast cancer risk with inconsistent results. Whether the different forms play different roles in breast cancer risk prediction is unclear. To examine this, we measured total and high molecular weight (HMW) adiponectin in a case-control study (1167 sets). Higher circulating HMW adiponectin was negatively associated with breast cancer risk after adjusting for menopausal status and family history of breast cancer (P=0.024). We analyzed the relationship between adiponectin and breast cancer risk in 6 subgroups. Higher circulating HMW adiponectin was also negatively associated with breast cancer risk (P=0.020, 0.014, 0.035) in the subgroups of postmenopausal women, negative family history of breast cancer, BMI>=24.0. Total adiponectin was positively associated with breast cancer (P=0.028) in the subgroup of BMI<=24.0. Higher HMW/total adiponectin ratio was negatively associated with breast cancer (P=0.019) in the subgroup of postmenopausal women. Interestingly, in the subgroup of women with family history of breast cancer, higher circulating total and HMW adiponectin were positively associated with breast cancer risk (P=0.034, 0.0116). This study showed different forms of circulating adiponectin levels might play different roles in breast cancer risk. A higher circulating HMW adiponectin is associated with a decreased breast cancer risk, especially in postmenopausal, without family history of breast cancer or BMI>=24.0 subgroups, whereas higher circulating HMW adiponectin levels is a risk factor in women with a family history of breast cancer. Further investigation of different forms of adiponectin on breast cancer risk is needed.