It has been widely acknowledged that anti-Müllerian hormone (AMH) is a golden marker of ovarian reserve. Declined ovarian reserve (DOR), based on experience from reproductive-aged women, refers to both the quantitative and qualitative reduction in oocytes. This view is challenged by a recent study clearly showing that the quality of oocytes is similar in young women undergoing IVF cycles irrespective of the level of AMH. However, it remains elusive whether AMH indicates oocyte quality in women with advanced age (WAA). The aim of this study was to investigate this issue. In the present study, we retrospectively analysed the data generated from a total of 492 IVF/ICSI cycles (from January 2017 to July 2020), and these IVF/ICSI cycles contributed 292 embryo transfer (ET) cycles (from June 2017 to September 2019, data of day 3 ET were included for analysis) in our reproductive centre. Based on the level of AMH, all patients (= > 37 years old) were divided into 2 groups: the AMH high (H) group and the AMH low (L) group. The parameters of in vitro embryo development and clinical outcomes were compared between the two groups. The results showed that women in the L group experienced severe DOR, as demonstrated by a higher rate of primary diagnosis of DOR, lower antral follicle count (AFC), higher level of basal follicle stimulating hormone (FSH) and cancelation cycles, lower level of E2 production on the day of surge, and fewer oocytes and MII oocytes retrieved. Compared with women in the H group, women in the L group showed slightly reduced top embryo formation rate but a similar normal fertilization rate and blastocyst formation rate. More importantly, we found that the rates of implantation, spontaneous miscarriage and livebirth were similar between the two groups, while the pregnancy rate was significantly reduced in the L group compared with the H group. Further analysis indicated that the higher pregnancy rate of women in the H group may be due to more top embryos transferred per cycle. Due to an extremely low implantation potential for transfer of non-top embryos from WAA (= > 37 years old) in our reproductive centre, we assumed that all the embryos that implanted may result from the transfer of top embryos. Based on this observation, we found that the ratio of embryos that successfully implanted or eventually led to a livebirth to top embryos transferred was similar between the H and the L groups. Furthermore, women with clinical pregnancy or livebirth in the H or L group did not show a higher level of serum AMH but were younger than women with non-pregnancy or non-livebirth. Taken together, this study showed that AMH had a limited role in predicting in vitro embryo developmental potential and had no role in predicting the in vivo embryo developmental potential, suggesting that in WAA, AMH should not be used as a marker of oocyte quality. This study supports the view that the accumulation of top embryos via multiple oocyte retrieval times is a good strategy for the treatment of WAA.