Abstract Background BMI has been implicated as a risk factor for heart disease as a whole in multiple studies. Heart attack is one of the common complications of this disease. The aim of this study is to explore if elevated level of BMI causes an increase in the risk of heart attacks. Methods We used two Mendelian randomisation (MR) methods: inverse variance weighted estimation and robust adjusted profile score (RAPS) on the basis of summary data of adulthood BMI from Genetic Investigation of Anthropometric Traits consortium and heart attack data from the UK Biobank. BMI associated single nucleotide polymorphisms (SNPs) were used as instrumental variables. Results Seventy-two independent SNPs were associated with BMI (P < 5 × 10^{− 8}). Using these SNPs as instruments, BMI was found to be causally associated with heart attacks in inverse variance weighted MR analysis. The risk of heart attacks increased by 0.8% per 1-SD (or 4.5 kg/m²) increase in BMI (OR = 1.008 with 95% CI (1.003, 1.012), P = 0.001). RAPS provided concordant results (OR = 1.007 with 95% CI (1.002, 1.012), P = 0.004). Conclusions This current study is the first to use MR to investigate causal relationship between BMI and heart attacks. Our findings suggest that high level of BMI may cause increased risk of heart attacks.