To assess which is the optimal protocol in terms of endometrial preparation prior to frozen-thawed embryo transfer (FET) in women with polycystic ovarian syndrome (PCOS) and to explore the effect in stimulated cycle with the addition of vaginal 17-beta oestradiol. Five hundred and seventy-six patients with PCOS were prepared for FET using artificial cycle induced with oestradiol and progesterone supplementation (n = 291) and stimulated cycle induced by human menopausal gonadotrophin (HMG) within or without the addition of vaginal 17-beta oestradiol (n = 285). Then the FET was performed in a receptive endometrium. Endometrial thickness was similar (9.03 +/- A 1.65 vs. 9.12 +/- A 1.58, P > 0.05) in artificial and stimulated cycle. The two protocols resulted in clinical pregnancy rate (41.0 % vs. 41.6 %, P > 0.05), ongoing pregnancy rate (36.6 % vs. 34.7 %, P > 0.05), live birth rate (30.0 % vs. 31.7 %, P > 0.05), which were not statistically different. Nevertheless, the cancelled cycle rate made a significant difference (2.2 % vs. 5.4 %, P < 0.05). There is no significant difference in the clinical pregnancy rate in HMG, HMG added with vaginal oestradiol and HMG switch to vaginal oestradiol group (42.6 %, 41.1 %, and 33.3 %, respectively). The mean endometrial thickness, clinical pregnancy rate, ongoing pregnancy rate, live birth rate and implantation rate were similar in artificial and stimulated cycle for endometrial preparation prior to FET in PCOS. It was fine to add vaginal 17-beta oestradiol to stimulated cycle when necessary. However, stimulated cycles had a significantly higher cancelled cycle rate. We should follow the principles of individualization, securitization and optimization in endometrial preparation of the FET in patients with PCOS.