BACKGROUND: Ovarian hyperstimulation syndrome (OHSS) is an iatrogenic and potentially life threatening condition resulting from excessive ovarian stimulation. Reported incidence of moderate to severe OHSS ranges from 0.6% to 5% of in vitro fertilization (IVF) cycles. The factors contributing to OHSS have not been completely explained. The release of vasoactive substances secreted by the ovaries under human chorionic gonadotrophin (hCG) stimulation may play a key role in triggering this syndrome. This condition is characterised by a massive shift of fluid from the intravascular compartment to the third space, resulting in profound intravascular depletion and haemoconcentration. OBJECTIVES: To assess the effect of withholding gonadotrophins (coasting) on the prevention of ovarian hyperstimulation syndrome in assisted reproduction cycles. SEARCH METHODS: For the update of this review, we searched the Cochrane Gynaecology and Fertility Group Trials Register, CENTRAL, MEDLINE (PubMed), CINHAL, PsycINFO, Embase, Google, and clinicaltrials.gov to 6 July 2016. SELECTION CRITERIA: We included only randomized controlled trials (RCTs) in which coasting was used to prevent OHSS. DATA COLLECTION AND ANALYSIS: Two review authors independently selected trials and extracted data. They resolved disagreements by discussion. They contacted study authors to request additional information or missing data. The intervention comparisons were coasting versus no coasting, coasting versus early unilateral follicular aspiration (EUFA), coasting versus gonadotrophin releasing hormone antagonist (antagonist), coasting versus follicle stimulating hormone administration at the time of hCG trigger (FSH co‐trigger), and coasting versus cabergoline. We performed statistical analysis in accordance with Cochrane guidelines. Our primary outcomes were moderate or severe OHSS and live birth. MAIN RESULTS: We included eight RCTs (702 women at high risk of developing OHSS). The quality of evidence was low or very low. The main limitations were failure to report live birth, risk of bias due to lack of information about study methods, and imprecision due to low event rates and lack of data. Four of the studies were published only as abstracts, and provided limited data. Coasting versus no coasting Rates of OHSS were lower in the coasting group (OR 0.11, 95% CI 0.05 to 0.24; I² = 0%, two RCTs; 207 women; low‐quality evidence), suggesting that if 45% of women developed moderate or severe OHSS without coasting, between 4% and 17% of women would develop it with coasting. There were too few data to determine whether there was a difference between the groups in rates of live birth (OR 0.48, 95% CI 0.14 to 1.62; one RCT; 68 women; very low‐quality evidence), clinical pregnancy (OR 0.82, 95% CI 0.46 to 1.44; I² = 0%; two RCTs; 207 women; low‐quality evidence), multiple pregnancy (OR 0.31, 95% CI 0.12 to 0.81; one RCT; 139 women; low‐quality evidence), or miscarriage (OR 0.85, 95% CI 0.25 to 2.86; I² = 0%; two RCTs; 207 women; very low‐quality evidence). Coasting versus EUFA There were too few data to determine whether there was a difference between the groups in rates of OHSS (OR 0.98, 95% CI 0.34 to 2.85; I² = 0%; 2 RCTs; 83 women; very low‐quality evidence), or clinical pregnancy (OR 0.67, 95% CI 0.25 to 1.79; I² = 0%; 2 RCTs; 83 women; very low‐quality evidence); no studies reported live birth, multiple pregnancy, or miscarriage. Coasting versus antagonist One RCT (190 women) reported this comparison, and no events of OHSS occurred in either arm. There were too few data to determine whether there was a difference between the groups in clinical pregnancy rates (OR 0.74, 95% CI 0.42 to 1.31; one RCT; 190 women; low‐quality evidence), or multiple pregnancy rates (OR 1.00, 95% CI 0.43 to 2.32; one RCT; 98 women; very low‐quality evidence); the study did not report live birth or miscarriage. Coasting versus FSH co‐trigger Rates of OHSS were higher in the coasting group (OR 43.74, 95% CI 2.54 to 754.58; one RCT; 102 women; very low‐quality evidence), with 15 events in the coasting arm and none in the FSH co‐trigger arm. There were too few data to determine whether there was a difference between the groups in clinical pregnancy rates (OR 0.92, 95% CI 0.43 to 2.10; one RCT; 102 women; low‐quality evidence). This study did not report data suitable for analysis on live birth, multiple pregnancy, or miscarriage, but stated that there was no significant difference between the groups. Coasting versus cabergoline There were too few data to determine whether there was a difference between the groups in rates of OHSS (OR 1.98, 95% CI 0.09 to 5.68; P = 0.20; I² = 72%; two RCTs; 120 women; very low‐quality evidence), with 11 events in the coasting arm and six in the cabergoline arm. The evidence suggested that coasting was associated with lower rates of clinical pregnancy (OR 0.38, 95% CI 0.16 to 0.88; P = 0.02; I² =0%; two RCTs; 120 women; very low‐quality evidence), but there were only 33 events altogether. These studies did not report data suitable for analysis on live birth, multiple pregnancy, or miscarriage. AUTHORS' CONCLUSIONS: There was low‐quality evidence to suggest that coasting reduced rates of moderate or severe OHSS more than no coasting. There was no evidence to suggest that coasting was more beneficial than other interventions, except that there was very low‐quality evidence from a single small study to suggest that using FSH co‐trigger at the time of HCG administration may be better at reducing the risk of OHSS than coasting. There were too few data to determine clearly whether there was a difference between the groups for any other outcomes.