Abstract Study question What is the impact of pregnancy on the morphological features and behaviour of ovarian endometrioma and deep endometriotic nodules? Summary answer For the majority of women, despite features of decidualization being common in the first and second trimesters, endometrioma and deep nodules will regress during pregnancy. What is known already Deep endometriosis and endometrioma subtypes are thought to affect approximately 5% of women in pregnancy, with about 50% being unaware of their condition. Pregnancy has a major effect on the size and morphological features of endometrioma, with published studies reporting a tendency for cyst regression. Decidualization, a hormonally induced pregnancy-related phenomenon, effects endometriomas and may raise suspicion of an ovarian malignancy. The behaviour of deep endometriosis in pregnancy is poorly understood and there is limited available literature on the subject. Study design, size, duration This was a prospective observational cohort study conducted over three years at a single centre. We included 65 women with a viable eutopic pregnancy and concomitant ultrasound features of deep and/or ovarian endometriosis. The study was conducted at University College London Hospital, which is a tertiary level referral unit for early pregnancy complications and an accredited Endometriosis Centre. Participants/materials, setting, methods All women who participated provided written consent and were invited for surveillance ultrasound examinations at the time of their routine scans in pregnancy. All scans were performed by a single operator to minimise interobserver error. The change in size of endometrioma and nodules were reported as change in their mean diameter. Endometrioma with irregular thick inner walls, hyperechoic papillary projections and/or high vascularity and hyperechoic nodules with moderate to high vascularity were reported as decidualized. Main results and the role of chance Sixty five women were included in the study. Their median age was 34 years (23-44), and the gestation at presentation was 7 + 6 weeks (3 + 6 to 18 + 0). 47/65(72%) were nulliparous, 48/65(74%) had a background of endometriosis and 19/65(29%) conceived following IVF. There were 10/65(15%, 95%CI 7-24) women with endometrioma alone, 28/65(43%, 95%CI 31-55) with nodules alone and the remaining 27/65(42%, 95%CI 30-54) had both. 29/34(85%, 95%CI 73-97) women with endometrioma experienced cyst regression, 2/34(6%, 95%CI 0-14) experienced cyst growth and in 10/34(29%, 95%CI 14-45) there was complete resolution of all cysts. 43/51(84%, 95%CI 74-94) women with nodules experienced nodule regression, 2/51(4%, 95%CI 0-9) experienced nodule growth and in 4/51(8%, 95%CI 0-15) there was complete resolution of all nodules. 5/37(14%, 95%CI 3-25) women who attended postnatal follow-up, experienced complete resolution of all endometriotic lesions during pregnancy . In 10/34(29%, 95%CI 14-45) women with endometrioma and 27/51(53%, 95%CI 39-67) with nodules, a pattern of growth was observed in the first and second trimesters, which preceded regression in later pregnancy. Features of decidualization were observed in 17/34(50%, 95%CI 33-67) women with endometrioma, most commonly in the 1st trimester, and 25/51(49%, 95%CI 35-63) women with nodules, most commonly observed in the 2nd trimester. Limitations, reasons for caution The lack of extended follow-up fails to establish the long-term impact of pregnancy, lactation and postnatal contraception on the behaviour of endometriosis. This study relies on ultrasound alone for the detection of moderate/severe disease with no correlation with laparoscopy. Wider implications of the findings Sonographic changes of endometriosis in pregnancy are difficult to differentiate from characteristics of malignant lesions. Better understanding of the appearance of endometriosis in pregnancy is vital to reduce unnecessary surgical procedures, associated morbidity to mothers and babies and will help clinicians to counsel women regarding the significance of their condition. Trial registration number The study was registered on Research Registry (Unique identifying number: researchregistry4569).