Percutaneous coronary intervention (PCI) for chronic total coronary occlusions (CTO) improves clinical symptoms and quality of life. The longer-term safety of PCI compared to optimal medical therapy (OMT) remains uncertain. We sought to evaluate the long-term safety of PCI for CTO in a randomised trial as compared to OMT. A total of 396 patients with a symptomatic CTO were enrolled into a randomised, multicentre clinical trial comparing PCI and OMT. Half of the patients had a single CTO; the others had multivessel disease. Non-CTO lesions were treated prior to randomisation (2:1 ratio). During follow-up, crossover from OMT to PCI occurred in 7.3% (1 year) and 17.5% (3 years) of patients. At 3 years, the incidence of cardiovascular death or nonfatal myocardial infarction was not significantly different between the groups (OMT 3.7% vs PCI 6.2%; p=0.29). By per-protocol analysis, the difference remained non-significant (OMT 5.7% vs PCI 4.7%; p=0.67). Overall, major adverse cardiovascular events (MACE) were more frequent with OMT (OMT 21.2% vs PCI 11.2%), largely because of ischaemia-driven revascularisation. The rates of stroke or hospitalisation for bleeding were not different between the groups. At 3 years there was no difference in the rate of cardiovascular death or myocardial infarction between PCI or OMT among patients with a remaining single coronary CTO. The MACE rate was higher in the OMT group due largely to ischaemia-driven revascularisation. CTO PCI appears to be a safe option for patients with a single remaining significant coronary CTO. CinicalTrials.gov: NCT01760083.