In the present study, the association between the incidence of second cancer and initial treatment for primary breast cancer was investigated using the Osaka Cancer Registry in Japan. We selected 45,575 patients diagnosed with breast cancer between January 1975 and December 2003. Information on initial cancer treatment and second cancer was obtained from the Osaka Cancer Registry. Patients were classified according to initial treatment (chemo-, hormone, or radiotherapy, or no treatment). We calculated the incidence rate ratio of second cancers in patients classified by treatment for the first cancer. The effects of treatment adjusted for covariates on second cancers were examined using stratified analyses and a Cox proportional hazard model. The final number of 33,043 subjects had a mean duration of follow-up of 5.2±4.3 years, during which 1,857 second cancers were diagnosed. For hormone therapy, the incidence rate ratio (IRR) of all second cancers was 0.64 [95% confidence interval (CI), 0.58-0.70], and that of corpus uteri cancer was 3.04 (95% CI, 1.78-5.19). The multivariate analysis revealed that the IRR of corpus uteri cancer associated with hormone therapy was 2.53 (95% CI, 1.41‑4.55). The incidence rate of all second cancers associated with initial treatment was lower than that associated with no treatment. Only second corpus uteri cancer may be related to hormone therapy.