Gonadotrophin treatment in clomiphene citrate resistant polycystic ovarian syndrome (PCOS) patients, using either low-dose step-up or low-dose step-down protocols, is highly effective to achieve singleton live births. Concomitant use of gonadotrophin releasing hormone analogues (GnRHa), which will block the endogenous feedback for monofollicular development during the low-dose step-up protocol, should not be employed. It is more difficult to induce ovulation in patients with more 'severe' PCOS, characterized by obesity and insulin resistance. There is need for optimization of starting doses for both the low-dose step-up and step-down protocols. Such optimization will prevent hyperstimulation due to a starting dose far above the FSH threshold, as well as minimize the time-consuming low-dose increments by starting with a higher dose in women with augmented FSH threshold. External validation of reported models for prediction of FSH response is warranted for tailoring and optimizing treatment for everyday clinical practice. Although preliminary, the partial cessation of follicular development, along with regression leading to atresia, lends support to the LH ceiling theory, emphasizing the delicate balance and need for both FSH and LH in normal follicular development. Future well-designed randomized controlled trials will reveal whether IVF with or without in-vitro maturation of the oocytes will improve safety and efficacy compared with classical ovulation induction strategies.