A meta-analysis was conducted to investigate whether flexible gonadotrophin-releasing hormone (GnRH) antagonist administration according to follicular size would be more beneficial than starting on a fixed day. Only randomized controlled trials were included after a comprehensive search strategy. The data were combined for meta-analysis with RevMan software. Eleven trials were identified, but only four randomized controlled trials met the inclusion criteria. There was no statistically significant difference in pregnancy rate per woman randomized, although there was a trend towards a higher pregnancy rate with the fixed protocol, especially with delayed administration beyond day 8 [odds ratio (OR) 0.7, 95% confidence interval (CI) 0.45-1.1]. There was no premature LH surge in any participant in either protocol. However, there was a statistically significant reduction both in number of antagonist ampoules (OR -1.2 95%, CI -1.26 to -1.15) and amount of gonadotrophin (OR 95.5 IU, 95% CI 74.8-116.1) used in the flexible protocol. In conclusion, there was no statistically significant difference in pregnancy rate between flexible and fixed protocols. There was a statistically significant reduction in the amount of recombinant FSH with the flexible protocol.