Objective To review the evidence for ethnic differences in susceptibility to adverse drug reactions (ADRs) to cardiovascular drugs. Design Systematic review and meta-analysis. Data sources We searched Medline and Embase to March 2005. Reference lists of identified articles were hand searched for further relevant articles. Review methods Studies were eligible for inclusion if they included at least two ethnic groups and one or more ADRs. We excluded case reports and case series. Results 564 studies contained some description of ethnicity and an ADR, and 132 of them related to cardiovascular therapies. Twenty four studies provided data for ADRs for at least two ethnic groups and were therefore eligible for inclusion. In pooled analyses the relative risk of angio-oedema from angiotensin converting enzyme (ACE) inhibitors in black compared with non-black patients was 3.0 (95% confidence interval 2.5 to 3.7); the relative risk of cough from ACE inhibitors was 2.7 (1.6 to 4.5) in East Asian compared with white patients; and the relative risk of intracranial haemorrhage with thrombolytic therapy was 1.5 (1.2 to 1.9) in black compared with non-black patients. Conclusion Patients from different ethnic groups have different risks for important ADRs to cardiovascular drugs. Ethnic group may therefore be one determinant of harms of a given treatment in the individual patient, either because it acts as a surrogate measure of genetic make up or because cultural factors alter the risk. Data are sparse, and regulators should consider asking for better data before licensing.