Background: Endometriosis is a common chronic, complex, and progressive gynecological disease which mainly affects women of reproductive age. It is defined by the presence and growth of ectopic endometrial tissue which has the histological characteristics of the normal endometrium. Researchers agree endometriosis is likely to be polygenic and multifactorial, but the exact pathogenic mechanisms are still unclear. Each theory singularly fails to account for all forms of endometriosis, thereby indicating multifactorial mechanisms. Development can be divided into five basic processes: adhesion, invasion, recruiting, angiogenesis, and proliferation. Genetics, biomolecular aberrations in the eutopic endometrium, dysfunctional immune response, peritoneal environment may all ultimately be involved. The most common symptoms are dysmenorrhea, dyspareunia, and lower back pain that worsens during menses. Depending on the location of the implants, rectal pain and painful defecation may also occur. The diagnosis of endometriosis should be considered especially if a patient develops dysmenorrhea after years of pain-free menstrual cycles. Infertility may also be the presenting complaint. Clinically, endometriosis should be considered in any woman of reproductive age with pelvic pain or infertility. However, since symptoms are often nonspecific, the disease may be misdiagnosed as other clinical conditions characterized by chronic pelvic pain (irritable bowel syndrome, interstitial cystitis/painful bladder syndrome, recurrent cystitisoveractive bladder), thus leading to inadequate treatment and considerable diagnostic delay. The diagnosis of endometriosis is delicate and complex, the ‘gold standard’ for a definitive diagnosis is laparoscopy with biopsy and histological analysis. Laparoscopy is the invasive diagnostic procedure, therefore, huge number of clinical studies have been focused on detection of non-invasive diagnostic procedures that would facilitate the diagnosis. Aim: The main aim of this study was to examine whether the determination of vascular endothelial growth factor (VEGF), Neuropilin 1 and Annexin A2 concentrations in serum may contribute to the diagnosis of endometriosis. The diagnostic accuracy of VEGF, Neuropilin 1 and Annexin A2 in distinguishing between healthy subjects and patients with endometriosis was also evaluated. Participants and methods: A total of 200 women aged 18–39 years were included in the results of this study. Preoperative serum samples were obtained after written informed consent from 200 candidates for laparoscopic surgery due sterilisation and suspected endometriosis based on ultrasound vaginal probe, clinical examination, infertility, medical history. Our exclusion criteria for all groups were as follows: pregnancy before surgery or 6 months after surgery, active pelvic inflammatory disease, any anti-inflammatory or hormonal or immunomodulatory medications use in the preceding 3 months, genital or any extragenital malignant disease, polycystic ovaries, cysts, adenomyosis, fibroids, other defined causes of infertility, use of oral contraceptives or gonadotropin-releasing hormone agonist (GnRH) analogues, chronic or acute genital disease except endometriosis, history of any medical problem (e.g., endocrine abnormalities, gastrointestinal, cardiovascular, and pulmonary system diseases), any systemic diseases (e.g., chronic or acute inflammatory disease, autoimmune disorders, hepatic or renal insufficiency). Candidates were divided into groups (endometriosis and control). Endometriosis group was comprised of 120 patients with endometriosis diagnosed by laparoscopy or laparotomy and confirmed by histopathological examination. Finally, endometriosis group comprised 120 patients divided into 3 groups; the first group (stage II endometriosis group) comprised 50 patients; while the second group (stage III endometriosis group) comprised 62 patients and the third group comprised 8 patients diagnosed with stage IV endometriosis, respectively, according to rASRM classification. Eighty healthy women who were undergoing laparoscopy because of sterilization were the control group. Serum concentrations of VEGF, Neuropilin 1 and Annexin A2 were determined in 120 patients with endometriosis and 80 healthy subjects. Blood samples were taken from all women undergoing laparoscopy half an hour before the induction of anesthesia, for the collection of serum. Five mililitres of venous blood were aseptically collected and left to coagulate, after which samples were centrifuged for 15 min at 2000 g and the serum was separated in a sterile tube. One test tube for all subjects (for VEGF and ANNXA2) was centrifuged after 30 minutes, and the NRP1 tube was centrifuged two hours after were aseptically collected. Serum concentrations of all three analytes were assayed by a standardised sandwich enzyme-linked immunosorbent assay (ELISA). Results: Higher VEGF, Neuropilin 1 and Annexin A2 concentrations were found in patients with endometriosis. Significant differences were detected between control (VEGF= 0,214 (0,201 – 0,230) μg/L, p<0.0001; NRP1 = 8.2 (7.3 – 9.4) μg/L, p<0.0001; ANXA2 = 12,4 (10,2 – 13,4) μg/L, p<0.0001) and endometriosis groups (VEGF= 0,254 (0,231 – 0,268) μg/L, p<0.0001; NRP1 = 27,1 (25,6 – 28,1) μg/L, p<0.0001; ANXA2 = 29,1 (21,3 – 35,6) μg/L, p<0.0001) in terms of VEGF, NRP1 and ANXA2 levels. VEGF and Annexin A2 concentrations showed a statistically significant difference depending on the severity of the disease and a number of locations affected by endometriosis. There were no significant correlations between VEGF anf Neuropilin 1 also between Neuropilin 1 and Annexin A2 in patients with endometriosis. Annexin A2 were positively associated with VEGF (r=0,73; p<0,001) in patients with endometriosis. Patients with a positive family history had a shorter time period in diagnosing the disease. VEGF and Annexin A2 concentrations were higher in patients with endometriosis in whom a symptom of pain was present. The results of the ROC curves of all three analytes indicate that Neuropilin 1 (AUC = 0.97 at the cut-off value > 11 μg / L) and Annexin A2 (AUC = 0.99 at the cut-off value > 16 μg / L) showed excellent, while VEGF (AUC = 0.89 at a cut-off value > 0.225 μg / L) showed very good ability to distinguish between healthy and endometriosis patients. The model combining Neuropilin 1 and Annexin A2 demonstrated diagnostic performances (AUC (95%CI) = 0,96 (0,92 – 0,99)). The proposed model correctly identifies 98 % of patients with endometriosis. Conclusions: In complex diseases such as endometriosis, due to the complexity of the pathophysiological background, it is unlikely that a single marker will reflect the changes and predict the outcome of the disease. A diagnostically non-invasive marker of endometriosis is needed to make a diagnosis, guide targeted treatment, and monitor activity and therapeutic response at all stages of the disease. Since it cannot yet be concluded that all pathophysiological processes in the development of endometriosis are known, markers need constant reevaluation and retrospective studies to justify their use. Improved understanding of the molecular mechanisms of endometriosis pathogenesis involving VEGF, Neuropilin 1 and Annexin A2 could help identifying novel targets for diagnosis and therapeutic interventions in endometriosis. Diagnostic characteristics of the proposed model, obtained by combining serum concentrations of VEGF, Neuropilin 1 and Annexin A2 suggested potential value as a noninvasive approach in endometriosis diagnosis.