AIM: We describe the safety profile of nifedipine GITS as assessed from adverse events reported in the ACTION trial in which 7,665 patients with stable, symptomatic coronary artery disease were randomly assigned nifedipine GITS or placebo and followed for a mean of 4.9 years. METHODS: All adverse events were coded using the COSTART coding dictionary. The incidence rate for each event was calculated as the number of patients with the event concerned divided by the total time 'at risk'. Hazard ratios comparing nifedipine with placebo and their 95% confidence intervals were obtained by Cox proportional-hazards analysis. RESULTS: As reported previously, nifedipine significantly reduced the incidence of cardiovascular events and procedures [hazard ratio (HR) 0.89, 95% confidence interval (CI) 0.83-0.95]. Apart from the known side effects of nifedipine, which include peripheral oedema, vasodilatation, hypotension, asthenia, constipation, leg cramps, non-specific respiratory complaints, impotence and polyuria, and which were reported more frequently in patients assigned nifedipine, the incidence rates of most other adverse events were similar. There were no differences in the occurrence of gastrointestinal haemorrhage, myocardial infarction and suicide. The rate of occurrence of death or new cancer excluding non-melanoma skin cancer for patients with no history of cancer at baseline was 2.53/100 patient years for patients assigned nifedipine and 2.37/100 patient years for patients assigned placebo (HR 1.06, 95% CI 0.93-1.22). CONCLUSION: Overall nifedipine GITS was well tolerated by patients with stable symptomatic angina.