STUDY QUESTION: Does a GnRH agonist (GnRHa) trigger followed by a bolus of 1.500 IU hCG in a group of patients at risk of ovarian hyperstimulation syndrome (OHSS) reduce the OHSS incidence compared with hCG trigger? SUMMARY ANSWER: A GnRHa trigger followed by early luteal hCG support with one bolus of 1.500 IU hCG appears to reduce OHSS in patients at risk of OHSS; however, in a low-risk group a second bolus of 1.500 IU hCG induced two cases of late onset OHSS. WHAT IS KNOWN ALREADY: A GnRHa trigger is an alternative to hCG in GnRH antagonist co-treated cycles. STUDY DESIGN, SIZE, DURATION: Two RCTs were performed in four Danish IVF units. A total of 446 patients were assessed for eligibility and 390 patients were enrolled in the study from January 2009 until December 2011. The primary outcome of the study was OHSS incidence in the group at risk of OHSS. PARTICIPANTS/MATERIALS, SETTING, METHODS: Patients received a fixed dose of recombinant human FSH for the first 4 days. On the day of triggering, patients were assessed for their risk of OHSS based on the total number of follicles ≥11 mm diameter, and were classified as being at risk of OHSS when the total number of follicles ≥11 mm was between 15 and 25 and at low risk of OHSS when the total number of follicles ≥11 mm was ≤14. Two separate randomization lists were used for each of the OHSS risk groups. Women at risk of OHSS were allocated (RCT 1) to either: Group A (n = 60), ovulation triggering with a bolus of 0.5 mg buserelin (GnRHa) s.c. followed by a single bolus of 1.500 IU hCG s.c. after the oocyte retrieval-or: Group B (n = 58): 5.000 IU hCG. Similarly, women at low risk of OHSS were allocated (RCT 2) to receive either: Group C (n = 125), a bolus of 0.5 mg buserelin s.c., followed by a bolus of 1.500 IU hCG s.c. after oocyte retrieval and a second bolus of 1.500 IU hCG on the day of oocyte retrieval +5-or: Group D (n = 141), 5.000 IU hCG. Groups C and D were included in order to obtain preliminary data. MAIN RESULTS AND THE ROLE OF CHANCE: In women at risk of OHSS (RCT 1) (15-25 follicles) no OHSS case was seen in Group A (GnRHa trigger and one bolus of 1.500 IU hCG), whereas two cases of moderate late-onset OHSS occurred in group B (3.4%), (P = 0.24). In contrast, in women at a low risk of OHSS (RCT 2) (≤14 follicles) two cases of late-onset OHSS occurred in Group C (GnRHa trigger and two boluses of 1.500 IU hCG), whereas no OHSS case was encountered in Group D (P = 0.22). LIMITATIONS, REASONS FOR CAUTION: Although the first RCT was powered to include 168 patients at risk of OHSS (15-25 follicles ≥11 mm) randomized to either GnRHa trigger or hCG trigger, the trial was prematurely discontinued when a total of 118 patients at risk of OHSS were randomized. In addition the second RCT in the OHSS low-risk group was designed as a feasibility study to assess the incidence of OHSS after GnRHa trigger and dual hCG administration versus 5.000 IU hCG. No power calculation was performed for this trial. In addition, there was a lack of blinding in the RCTs. WIDER IMPLICATIONS OF THE FINDINGS: Although a non-significant result, one bolus of 1.500 IU hCG after GnRHa trigger tended to reduce the OHSS rate in patients with 15-25 follicles ≥11 mm as well as secure the ongoing pregnancy rate. In contrast, in patients at low risk of OHSS the administration of two boluses of 1.500 IU hCG after GnRHa trigger should be avoided as it may induce OHSS. STUDY FUNDING/POTENTIAL COMPETING INTERESTS: The study was supported by a research grant by MSD, Denmark. No conflict of interest was declared. TRIAL REGISTRATION NUMBER: Project number: VN-20060036MCH. CLIN TRIAL.GOV NUMBER: NCT00627406.