Endometriosis is an estrogen-dependent disease. Peritoneal endometriosis and endometriotic (adenomyotic) nodules of the recto-vaginal septum are distinct clinical entities with different histopathogeneses. The mechanisms and etiological factors involved in the onset or the development of these entities remain incompletely elucidated. Environmental pollutants (e.g. heavy metals and organochlorines) are among suspected etiological factors of endometriosis. Cadmium and lead are known to exhibit endocrine disrupting properties. Dioxins and PCBs have been shown to promote the development of endometriosis experimentally in primate and rodent models but the results of epidemiological studies remain inconclusive. We designed a matched case-control study to compare risk factors associated with endometriosis (E) and deep endometriotic nodules (DEN). Eighty eight triplets of women suffering from DEN, E or controls were recruited. Controls were not recruited in infertility clinics. Particular attention was given to dioxin exposure, which occurs essentially via the consumption of contaminated food. Dioxin exposure was assessed through an individual dietary questionnaire, and by GC/HRMS measurement in serum in a subset of patients who volunteered to give blood. The questionnaire revealed a significant protection associated with (multiple)-parity and breastfeeding in women with E or DEN, whereas the use of tampons or sanitary towels was not significantly associated with any form of the disease. Alcohol consumption was more prevalent (Odds Ratio (OR) 5.82 [CI95% 1.20-28.3]) in women suffering from DEN and a limited physical activity at work appeared as a risk factor for DEN and E (OR 4.58 [CI95% 1.80-11.62] and 5.61 [CI95% 1.90-16.60], respectively). Environmental or occupational exposures as well as dietary habits assessed through a food frequency questionnaire focusing on the consumption of fatty foodstuffs, were not significantly associated with the diseases. We measured the serum concentration of dioxins (pg TEQ/ g lipids) in 71 patients (E (n=25), DEN (n=25), and controls (n=21)). Age and body mass index were traced by linear multiple regression as determinants of dioxins while dietary habits and parity had no apparent influence. A logistic regression analysis was performed to take into account serum dioxins, age, body mass index, cumulative use of oral contraceptive, age at menarche, menstrual characteristics and familial factors. Significantly increased risks associated with dioxin serum concentration were found when comparing patients suffering from DEN and E with controls (OR 7.70 (CI95% [1.97-30.17]) and 3.17 (CI95% [1.01-9.93] respectively). This is the first epidemiological evidence of an association between organochlorines and endometriosis. We also measured the concentrations of heavy metals (cadmium and lead) in urine and/or blood in order to explore their possible role in endometriosis. Our observations did not support a role for cadmium in the onset or growth of endometriotic diseases but suggest a relationship with lead. Because endometriosis is an estrogen-dependent disease and since aromatase (CYP19), a key enzyme in estrogen biosynthesis, was recently demonstrated to be expressed within endometriotic lesions, we hypothesized that organochlorines could modulate local estrogen production through an up-regulation of aromatase. To explore this hypothesis, we firstly assessed the expression of aromatase in biopsied endometriotic tissue (peritoneal and ovarian endometriosis and deep endometriotic nodules of the recto-vaginal septum). First, we found that the expression of aromatase was significantly different in each type of tissue, which strengthens the theory of three distinct clinical entities. Compared with peritoneal endometriosis, ovarian endometriosis exhibited an 8-fold higher expression of aromatase, suggesting that aromatase inhibitors may be particularly active in this form of endometriosis. We then considered the concentration of polychlorobiphenyls, polychlorinated-dibenzo-dioxins and polychlorinated-dibenzo-furans in serum of these patients. The relationship between these parameters and aromatase expression was traced by simple and multiple regression analysis. The results did not support our hypothesis that the examined compounds facilitate the growth of endometriotic lesions by up-regulating the expression of aromatase. Overall, this work has contributed to better delineate the association between environmental exposures to organochlorines and the onset or development of peritoneal endometriosis or deep endometriotic nodules of the rectovaginal septum.