STUDY QUESTION: How does the frequency and location of endometriosis lesions observed at laparoscopy vary with patient age? SUMMARY ANSWER: We find that older patients have more deep lesions and ovarian endometriomas at laparoscopy, although this plateaus around 30 years of age. WHAT IS KNOWN ALREADY: The natural history and aetiology of endometriosis remain uncertain, with evidence of progression, spontaneous regression, and stable disease in various studies. STUDY DESIGN, SIZE, DURATION: Cross-sectional analysis of a prospectively collected surgical registry, including data from 104 accredited endometriosis centres across 7 countries, with 14 670 cases in total. PARTICIPANTS/MATERIALS, SETTING, METHODS: We analysed data from the British Society for Gynaecological Endoscopy (BSGE) endometriosis centres database, a multicentre international prospective cohort of patients undergoing surgery for deep endometriosis. We included all patients aged 10-55 years from 2009 to 2021. The change in lesion prevalence with age was modelled using three different logistic regression models-non-quadratic, quadratic, and segmented, adjusted for whether they had undergone endometriosis excision in the past. MAIN RESULTS AND THE ROLE OF CHANCE: We identified 14 670 cases from 104 centres across 7 countries. The mean age at the time of surgery was 35.9 years (SD 7.4). The odds of superficial endometriosis decreased linearly with age by 3% per year (adjusted odds ratio (aOR) 0.97, 95% CI 0.96-0.97, P < 0.001). The following increased before plateauing: bowel endometriosis increased by 7% per year (aOR 1.07, 95% CI 1.05-1.09, P < 0.001) until 32.9 years of age (95% CI 30.5-35.3); deep sidewall disease increased by 7% per year (aOR 1.07, 95% CI 1.05-1.09, P < 0.001) until 31.3 years (95% CI 29.2-33.5); and uterosacral disease increased by 4% per year (aOR 1.04, 95% CI 1.02-1.06, P < 0.001) until 33 years (95% CI 29.3-36.6). Deep bladder disease increased by 12% per year (aOR 1.12, 95% CI 1.04-1.22, P = 0.005) until 28.2 (95% CI 25.8-30.6) years, after which the odds decreased by 2% per year (aOR 0.98, 95% CI 0.97-1.00, P = 0.007). Ovarian endometrioma was best described by a quadratic model, with a starting increase in odds of endometrioma of 20% per year at 10 years (aOR 1.20, 95% CI 1.16-1.23, P < 0.001), and a peak prevalence at 40.9 years of age (95% CI 34.1-47.7, P < 0.001). The total number of sites affected by endometriosis increased by 0.12 sites per year (95% CI 0.10-0.14, P < 0.001) until 32.8 years of age, with no correlation with age after this point. LIMITATIONS, REASONS FOR CAUTION: This cross-sectional study only includes patients undergoing excision with pararectal dissection for the treatment of pain, so does not capture patients who do not undergo such surgery, including asymptomatic patients, and cannot provide data on potential disease progression on an individual level. WIDER IMPLICATIONS OF THE FINDINGS: This study supports existing evidence suggesting that older patients with endometriosis have more deep nodules and endometriomas and less superficial disease. STUDY FUNDING/COMPETING INTEREST(S): No funding was provided for this study. K.V. reports consulting fees from Eli Lilly, Gedeon Richter, Gesynta, and Rickett (all paid to institution); speakers' fees from Gedeon Richter and Reckitt (both paid to institution); membership of the Medical Advisory Panel of Endometriosis UK (unpaid); past presidency of the International Association for the Study of Pain (IASP) Special Interest Group (SIG) on Abdominal and Pelvic Pain (unpaid). TRIAL REGISTRATION NUMBER: N/A.