AIMS/HYPOTHESIS: Two US randomised trials found a lower incidence of type 2 diabetes in women treated by menopausal hormone therapy (MHT) with oral conjugated equine oestrogen combined with medroxyprogesterone acetate. The purpose of this study was to evaluate the influence of various MHTs, according to their formulation and route of administration, on new-onset diabetes in a cohort of postmenopausal French women. METHODS: The association between MHT use and new-onset diabetes was investigated by Cox regression analysis in 63,624 postmenopausal women in the prospective French cohort of the Etude Epidemiologique de Femmes de la Mutuelle Générale de l'Education Nationale (E3N). Cases of diabetes were identified through self-reporting or drug-reimbursement record linkage, and further validated. RESULTS: 1,220 new-onset diabetes cases were validated. We observed a lower risk of new-onset diabetes among women who had ever used MHT (HR 0.82 [95% CI 0.72-0.93]), compared with those who had never used MHT. Adjustment for BMI during follow-up (rather than according to baseline BMI) did not substantially modify this association. An oral route of oestrogen administration was associated with a greater decrease in diabetes risk than a cutaneous route (HR 0.68 [95% CI 0.55-0.85] vs 0.87 [95% CI 0.75-1.00], p for homogeneity = 0.028). We were not able to show significant differences between the progestagens used in combined MHT. CONCLUSIONS/INTERPRETATION: Use of MHT appeared to be associated with a lower risk of new-onset diabetes. This relationship was not mediated by changes in BMI. Further studies are needed to confirm the stronger effect of oral administration of oestrogen compared with cutaneous administration.