The response to ovarian stimulation is the key determinant of ART success, but there are many definitions of poor response and many proposed solutions. The response to ovarian stimulation is the key determinant of ART success, but there are many definitions of poor response and many proposed solutions. The response to ovarian stimulation is a key and major determinant of the outcome of treatment with assisted reproductive technique. Age, antral follicle count, and anti-müllerian hormone level seem to be the best indicators to predict the number (but not the quality) of the oocytes that will be collected. This suggests that most cases of inadequate response to ovarian stimulation can be predicted by a careful appreciation of ovarian reserve. Certainly, there are many unexpected instances of poor ovarian response in young women. In all such cases, we must control the ovarian reserve to make sure that medications have been well administered with respect to weight and other medical history that could interfere with ovarian stimulation. One can decide to modify the protocol, perhaps after assessing polymorphisms in the genes of gonadotropin receptors. Even in “predictable” poor ovarian response, the definition of poor response is a huge problem, leading to difficulties interpreting the results of research protocols. As Paul Devroey will show in a contribution here, there are more than 40 definitions of poor response. The recently published criteria proposed by the Bologna ESHRE meeting are not absent of criticism especially because we do not know the various mechanisms of diminishing oocyte quantity, which can have different impacts on oocyte quality and may not be sufficient to improve the outcomes of various treatments. Findings from very few randomized controlled trials apply to a clearly defined group of patients. What kind of treatment can we propose in this situation? Many controversial propositions can be considered, such as increasing the dose of gonadotropins, prolonged administration of gonadotropins, the use of flare-up and micro flare-up, adjunct use of LH, homogenization of the cohort by E2 in a previous luteal phase, delayed administration of LH, or also to change agonist into antagonist blockage for LH surge. Several authors have proposed different adjuvants like l-arginine, dexamethasone, colony-stimulating factor–1, pyridostigmine, aspirin, or other biological techniques. Three approaches presented in the following commentaries are more likely to be used and of particular interest:1.Kadoch comments on the natural cycle, mainly with the following question: What is the point of trying to stimulate so hard to end up collecting a maximum of two embryos when repeated natural cycles can be realized?2.Growth hormone cotreatment in case of poor COS response has been proposed. De Ziegler and colleagues identified 139 publications over the past 5 years and 3 recent meta-analyses, and they comment on the potential effect on oocyte quality even when GH supplements did not increase the quantity of oocytes retrieved.3.The third proposition that will be analyzed by Fanchin is the use of androgenic supplementation. Based on the excess of small antral follicle number in androgenic diseases such as polycystic ovary syndrome, and the observation of artificial polycystic ovary syndrome in induced transsexual women who want to virilize themselves, different androgenic supplementation by different routes and models have been proposed. Another approach discussed is the use of a small amount of LH to increase the size, number, and perhaps the quality of retrieved oocytes. Certainly, after these interesting contributions, physicians will have the conviction that further trials are urgently needed to define subgroups of poor prognosis with assisted reproductive technique and predict poor response in younger patients. Only this classification will enable us to compare results adequately.