OBJECTIVES: The aim of this work was to verify the existence of and characterise the relationship between premenopausal progestogen use, subsequent use of estrogen-progestogen therapy (EPT) and risk factors for breast cancer, and to describe and analyse premenopausal progestogen use among women participating in a longitudinal study. METHODS: Data came from self-administered questionnaires mailed to 2254 pre- or peri-menopausal women aged 45 years or older participating in the French GAZEL cohort and followed for 10 years. Bivariate and multivariate analyses first examined the association between progestogen use and women’s and physicians’ characteristics. A generalised linear model was then used to analyse the association between progestogen use and EPT use, while taking into account factors associated with progestogen use. RESULTS: Thirty-six percent of the women had used progestogens before menopause: 56% of them 19-norpregnane derivatives, and 13% 17alpha hydroxyprogesterone derivatives. They were more likely to report breast pain, a family history of breast cancer and mood changes, to be thin, to consult a gynaecologist rather than a general practitioner, to consult a female physician, and to consult often. Of the 1756 women who had reached menopause, EPT users were more likely to have used progestogens premenopausally (adjusted RR: 1.2; 95% CI: 1.1-1.3). CONCLUSIONS: Physician characteristics play an important role in the use of progestogens before menopause. Our results also suggest that women using progestogens may have a higher risk of breast cancer, that is, that progestogen use may be a confounding rather than causal factor. The analyses concerning the relationship between EPT use and breast cancer risk must take progestogen use before menopause into account, as they do other confounding factors, such as age at menarche, parity, and use of oral contraceptives.