It is hypothesized that insulin resistance and related metabolic factors may influence breast cancer risk, however the epidemiological evidence remains inconclusive. We conducted a case–control study nested in a prospective cohort in Northern Sweden, to clarify the associations of body mass index (BMI), leptin, adiponectin, C-peptide, and glycated haemoglobin (HbA1c) with breast cancer risk. We also investigated whether these associations may be modified by age at diagnosis, tumour stage, and oestrogen and progesterone receptor status. During follow-up, 561 women developed invasive breast cancer and 561 matched controls were selected. Conditional logistic regression was used to calculate odds ratios (OR) as estimates of relative risk, and 95% confidence intervals (CI). The associations of BMI, leptin and HbA1c with breast cancer risk differed significantly according to whether the tumour was diagnosed as stage I or stage II–IV (P heterogeneity all <0.05). These factors were significantly inversely associated with risk in the group of stage I tumours, with ORs for top vs. bottom tertile for BMI of 0.48 (95% CI, 0.30–0.78, P trend = 0.004); leptin, 0.64 (95% CI, 0.41–1.00, P trend = 0.06); and HbA1c, 0.47 (95% CI, 0.28–0.80, P trend = 0.005). For stage II–IV tumours, there was a suggestion of an increased risk with higher levels of these factors. There were no significant differences in the associations of BMI, leptin, adiponectin, C-peptide and HbA1c with breast cancer risk in subgroups of age at diagnosis or tumour receptor status. This prospective study suggests that BMI, leptin and HbA1c influence breast tumour initiation and progression.