Objective: In the present study, human menstrual endometrium was transplanted into 66 nude mice in order to 1) examine the early stages of evolution of the endometriotic lesion, 2) to test whether the cellular, seric and red blood cell fractions from the menstrual effluent may have an influence on the development of endometriotic lesions in this animal model and 3) to investigate whether adhesion occurs on intact peritoneum or at the site of incision. Design: In a first series of experiments (n = 10), the timing of adhesion of menstrual endometrial tissue was analyzed by removal of the transplant on days 1, 3 and 5. In a second series (n = 25), 5 different fractions of menstrual effluent were transplanted by minilaparotomy performed on the ventral midline: 1) effluent as a whole, 2) seric fraction, 3) cellular + seric fractions 4) cellular + red blood cell fractions and 5) cellular fraction only. In a third series (n = 31), the same fractions, except the seric, were deposited by an incision on the ventral right side in order to test if the site of adhesion is influenced by the healing area. Materials/Methods: Menstrual effluent was obtained from women on day 2 or 3 of the cycle, by aspiration. The effluent was centrifuged and the seric fraction collected. Endometrial tissue was separated from red blood cells following two centifugations on Ficoll-Paque gradients and both fractions were resuspended in serum-free medium. Endometrial cell viability was checked by trypan blue exclusion. Minilaparotomy was performed in order to inject the various fractions into the peritoneal cavity of nude mice. Removal of the transplants was performed by laparotomy. Adhesion of endometrial cells and the early stages of development of endometrial lesions were studied morphologically and immunohistochemically. Staining of iron deposits and Ki67, a marker of cell proliferative activity, was carried out. The χ2 test was used for statistical analysis. Results: As early as day 1, stromal cells were found to adhere to the peritoneal mesothelium. A progressive reorganization of epithelial and stromal cells was observed. On day 5, typical endometriotic lesions, still with some necrotic areas but also large areas of normal-looking stroma and cystic glands with flattened or columnar epithelium, were present in 75% of cases. An increased rate was found when either total menstrual effluent or the cellular fraction associated with the red blood cell fraction was injected (82% and 89% respectively), when compared with rates observed after injection of the cellular fraction alone (50%) or associated with the soluble fraction (75%). A significant increase (p < 0.02) in glandular proliferative activity was observed on day 5 compared to values observed on days 1 and 3. The proliferative activity was similar whatever the injected fraction. Iron deposits were observed independently of the fraction injected. There was a tendency towards increased hemosiderin deposits when red blood cells were injected together with the cellular fraction. Implantation occurred on incision sites as well as intact mesothelium whether laparotomy was performed medially or on the right side. Conclusions: Implantation of endometrial cells can occur on intact mesothelium but may be favored by the presence of red blood cells as well as the occurrence of a peritoneal injury. Supported By: Grant 7.4530.98 from the “Fonds National de la PRecherche Scientifique de Belgique.”