Objective To quantify any risk of atrial fibrillation (AF) associated with ivabradine treatment by meta-analysis of clinical trial data. Methods Medline, Embase, Web of Knowledge and the Cochrane central register of controlled trials were searched for double-blinded randomised controlled trials of ivabradine with a minimum follow-up period of 4 weeks. For studies where AF data were unpublished, safety data were obtained from the European Medicines Agency (EMeA) website and personal communications. Studies were appraised for risk of bias using components recommended by the Cochrane Collaboration. Meta-analyses were performed of relative risk of AF and absolute risk difference of AF per year of treatment. The main outcome measure was incident AF during the follow-up period. Results AF data were available from 11 studies: one from the published report, six from the EMeA and four from personal communications. Ivabradine treatment was associated with a relative risk of AF of 1.15 (95% CI 1.07 to 1.24, p=0.0027) among 21 571 patients in the meta-analysis. From this we estimated that the number needed to harm for ivabradine would be 208 (95% CI 122 to 667) per year of treatment. Conclusions AF is a substantially more common side effect of ivabradine treatment than one patient in 10 000, the risk presently reported in the product literature. The incidence of AF has not routinely been reported in clinical trials of ivabradine.