Abstract Study question Underlying mechanisms and specific treatment for endometriosis-related infertility are still unclear and lacking. Is there any animal model suitable for the study? Summary answer Endometium ligation to ovary fat pad in rats is a more appropriate and successful but less detrimental model of endometrioma for infertility pathology and outcomes. What is known already Nonhuman primates (NHP) have been the most representative animal model for endometriosis since they menstruate in a cyclic pattern and develop endometriosis spontaneously as in human. However, the incidence of spontaneous endometriosis in NHP is low and due to ethical concerns and high cost, the application of NHP for endometriosis study is restricted. Rodents (i.e. rats, mice, rabbits) have been used to analyze the pathophysiology of endometriosis. Because rodents do not develop endometriosis spontaneously, it has to be induced surgically. Ovarian endometriosis is the most common type, but previous experimental models were mostly either subcutaneous, peritoneal wall or mesenteric endometriosis. Study design, size, duration An animal study to compare different methods of endometrium transplantation to the ovary as potential animal model of endometrioma to study the endometriosis-related infertility. Compared with NHP and other rodents, rats were chosen due to easy access, low cost, good ovary size, short estrus and reproductive cycle and similar endocrine pattern to human beings. For each transplantation method, at least 5 animals were included and followed up in different time points for comparisons. Participants/materials, setting, methods To establish ovarian endometriosis or endometrioma model, uterine tissues were collected from donor rats and then transplanted to recipient rats by either adhesion to ovary, ligation to ovary fat pad, or injection underneath tunica albuginea. Vasculature and histology of engrafts, follicle count and morphology of ovary, receptivity markers of endometrium, immune response and inflammatory cytokines of peritoneal cavity and hormonal changes were assessed after 4 days of transplantation. Implantation rates after conception will be examined. Main results and the role of chance Compared with other transplantation methods, only ligation of endometrium fragments to ovary fat pad resulted in endometrioma-like lesion in the ovaries. Compare with sham control ovaries, formation of new vessels from surrounding ovarian vessels to the endometriotic engraft assessed by Cellvizio LAB in vivo imaging was identified, indicating active angiogenesis. Morphology and histology of endometriotic cyst were confirmed in the lesion by H&E staining, suggesting functional endometrium. Stroma markers CD140b, CD106 lkand epithelium markers keratin 17/19 and EpCam were found in the ovary cortex, implying integration of endometrium tissue to the ovarian tissues. Decreased AMH expression and antral follicle count in the ovaries suggested defective follicle development. Reduced receptivity markers HOXA10, LIF and ανβ3 integrin expression in uterine endometrium and implantation rate after nature mating or embryo transfer indicated potential impaired endometrial receptivity. Peripheral and peritoneal levels of COX2 and IL–8 were elevated, suggesting inflammatory response. However, estrogen and progesterone levels were not significant different from baseline and other transplantation methods. Limitations, reasons for caution This was an animal model, it might not totally reflect the exact pathological changes of endometrioma in human. Current method was a single and short-term transplantation model, monthly endometrial cells engraftment and slow growing lesions without sufficient estrogen supply might limit the establishment of endometrium-like lesions. Further testings are needed. Wider implications of the findings: Establishment of animal models for endometrioma is vital important to investigate the pathophysiology, underlying mechanism and potential therapy targets for endometriosis-related infertility. Currently such model is still lacking. An appropriate, adequate and effective but less detrimental model of endometrioma can accelerate the scientific research and clinical application in near future. Trial registration number N/A