Abstract Background Thyroid function has been associated with breast cancer risk, and breast cancer cell growth and proliferation. It is not clear whether thyroid function affects prognosis following breast cancer but, if so, this could have an important clinical impact. The present study analysed prospectively collected measurements of free tri-iodothyronine (T3), free thyroxine (T4), thyroid-stimulating hormone (TSH) and thyroid peroxidase antibodies (TPO-Ab) in relation to breast cancer survival. Methods The Malmö Diet and Cancer Study is a prospective cohort study of 17 035 women in Sweden. Study enrolment was conducted between 1991 and 1996. Patients with incident breast cancer were identified through record linkage with cancer registries until 31 December 2006. Information on vital status was collected from the Swedish Cause of Death Registry, with the endpoint breast cancer mortality (31 December 2013). Hazard ratios (HRs) with 95 per cent confidence intervals (c.i.) were obtained by Cox proportional hazards analysis. Results Some 766 patients with incident breast cancer were identified, of whom 551 were eligible for analysis. Compared with patients in the first free T4 tertile, breast cancer mortality was lower among those in the second tertile (HR 0·49, 95 per cent c.i. 0·28 to 0·84). There was an indication, although non-significant, of lower breast cancer mortality among patients in the second TSH tertile (HR 0·63, 0·37 to 1·09) and in those with positive TPO-Ab status (HR 0·61, 0·30 to 1·23). Free T3 showed no clear association with mortality. Conclusion In the present study, there was a positive association between free T4 levels and improved breast cancer survival.