Since several years, recombinant FSH (r-FSH) have been replacing urinary products in controlled ovarian hyperstimulation (COH), particularly in ART, because of hypothezised advantages on safety and efficacy, which resulted in a large increase of cycles cost. However, several recent meta-analyses were unable to find any difference in terms of pregnancy rate (PR), and no specific safety risk was ever demonstrated for urinary products. Moreover, the beneficial role of LH activity has been analyzed in many studies, leading to use more widely hMG. In this debate, it was useful to better understand the reasons related to the medical choice of one type of drug or the other. French multicenter study carried in 20 voluntary centers. Retrospective analysis of clinical data routinely collected in their medical files in 2003 and 2004 . 20 centers were recruited on the following criteria: routine use of hMG (more than 10% of the cycles), willingness to participate, existence of computerized routine data on IVF. A form was designed, including the patients’ age, the infertility characteristics (infertility duration, past history of pregnancy, ovulation quality, tubal status, endometriosis, semen characteristics), the main cycle characteristics (down regulation, controlled ovarian hyperstimulation) oocyte recovery, laboratory results and cycle outcome. Two groups were constituted, one with cycles treated with hMG, and the second with r-FSH. At this step, patients receiving both types of drugs or with spontaneous cycles were excluded. Those two groups were compared on all the patients’ characteristics, to analyze the factors implicated in the medical choice, but not on the efficacy results, since the hypothesis was that patients were not similar and the drug administration was not randomized. The statistical analysis was done using the SAS software, with tests adapted to the studied variables nature, and a p value at 0.05 for significance. The final population of the study was composed of 15,000 cycles, of which 24% received hMG. Women treated with hMG represented 27.5% of IVF cycles and 22.1% of ICSI cycles. The group receiving hMG was constituted of women older than the other (35.7 [SD=4.4] vs. 34.2 [4.5], p<0.001). There were more women aged more than 35 years (60.4 % vs. 45.4%, p<0.001), and aged more than 40 (22.9% vs. 13.3%, p<0.001). Men were also older, even if the difference was smaller (37.0 [5.8] vs. 36.3 [5.8], p<0.01). There was no difference in infertility duration (5.6 [2.7] in hMG group vs. 6.5 [3.1], p=0.59). The number of previous ART cycles was higher in the hMG group (1.3 [1.5] vs 1.0 [1.3]. In conventional IVF, there were more blocked tubes (42.7% vs. 36.4%, p<0.01), more endometriosis (7.2% vs. 5.0%, p=0.02). The percentage of male infertility and of unexplained infertility were similar (respectively 44.9% vs. 45.8%, p=0.64 and 14.0% vs. 11.6%, p=0.09). In patients treated by ICSI, those without male infertility were more frequent in the r-FSH group (19.8% vs. 12.1%, p<0.001). Clearly, the patients’ characteristics have an influence on the medical choice of the drug used for COH. This must be carefully taken in account when comparing the results of those products in non randomized observational studies.