Abstract Study question Are circulating levels of molecular forms of AMH throughout pregnancy still higher in women with PCOS than in controls? Summary answer AMH and its molecular forms decrease significantly throughout pregnancy, remaining higher in women with PCOS, without modification in the proportion of AMH molecular forms. What is known already AMH injection in late gestational mice causes a “PCOS” phenotype in the offspring’s, possibly via a gestational hyperandrogenism. AMH could therefore play a role in the heritability of PCOS in women. Circulating AMH exists in different molecular forms, cleaved (active) and uncleaved. In normal women, serum AMH decreases during pregnancy. In women with PCOS, cross-sectional or partial longitudinal studies have shown higher AMH levels in the second trimester of pregnancy and at the time of delivery in women with PCOS compared to controls. To date, no longitudinal dynamic monitoring of AMH throughout pregnancy in women with PCOS has been performed. Study design, size, duration Systematic prospective quarterly longitudinal monocentric comparative follow up of 30 women with PCOS and 29 controls before and during pregnancy from April 2019 to September 2021. To demonstrate a high effect size with 90% power and 5% first-species risk, it was necessary to include 23 subjects per arm. With a calculated loss of follow-up rate of 20%, 29 patients must be recruited per group. Participants/materials, setting, methods Women aged 18 to 43y, with a pre-conception dosage of AMH, were included during the first trimester of a singleton pregnancy. PCOS group was defined according to the modified Rotterdam criteria. In the control group, the patients had normal ovarian reserve. During the first, second and third trimester of pregnancy, all women were tested for serum AMH and its molecular forms with a Beckman automatic analyser, and for serum estradiol, LH and androgen levels. Main results and the role of chance Before pregnancy, patients with PCOS had higher levels of AMH (58.9 vs. 18.3pmol/L; p = <0.001), LH (5.9 vs. 3.3IU/L; p = 0.004) and total testosterone (TT) (0.35 vs. 0.29ng/ml; p = 0.019) than controls. Within each group, the levels of AMH and cleaved AMH were significantly lower in the third trimester than before pregnancy. The serum AMH level in the third trimester remained significantly higher in women with PCOS than in controls (18.4 vs. 6.5pmol/L; p = 0.002). At all times, the amount of cleaved AMH was higher in women with PCOS than in controls. The proportion of AMH molecular forms was not different between women with PCOS and controls and did not change throughout pregnancy. The LH level, mostly undetectable, was significantly lower in the third trimester than before pregnancy in both groups. In the first trimester, it remained higher in women with PCOS (0.6 vs 0.4IU/L; p = 0.018). In each group, the TT level was significantly higher in the third trimester than before pregnancy. TT level remained higher in women with PCOS than in control women in the first (0.76 vs 0.55ng/ml; p = 0.022) and second trimesters 0.82 vs. 0.61ng/ml; p = 0.013). In the third trimester, it was no longer significantly different (0.97 vs 0.64ng/ml; p = 0.069). Limitations, reasons for caution Only free testosterone is active, and only TT was measured. Additional SHBG and albumin measurements are needed. Our control population is a population of infertile women who have no ovarian problem but who have undergone, for most of them, ART treatments to obtain pregnancy with a possible impact on implantation. Wider implications of the findings Elevated serum AMH level could be instrumental in the modification of the in-utero androgenic environment that is believed to promote the transmission of the syndrome. The impact of AMH could be different depending on the clinical or metabolic phenotype of the patients and needs to be explored further. Trial registration number NCT03483792