What is known and objective Although many studies have examined medication adherence in patients with coronary artery disease (CAD), no prospective trial has compared medication adherence between patients treated with percutaneous coronary intervention (PCI) or with optimal medical therapy (OMT) in real life. This study sought to compare the adherence to evidence-based secondary preventive medications in patients with documented CAD treated with PCI and OMT, or OMT alone. Methods We evaluated adherence to statins, beta-blockers, and angiotensin converting enzyme inhibitors (ACEI) during a 6-month follow-up in 232 patients with documented CAD, comparing patients treated with PCI and those receiving medical therapy alone. Medication adherence was measured with reference to national reimbursement database records. Results and discussion Of the 232 patients who survived the 6-month follow-up, the percentages of adherent patients according to prescription records (prespecified primary endpoint) were 53·6% (n = 82) in the PCI group and 33·8% (n = 27) in the OMT group (P = 0·004). Analysis of the individual medication classes revealed similar results for beta-blockers (86·0% in PCI group vs. 72·5% in OMT group, P = 0·006) and statins (64·5% in PCI group vs. 44·0% in OMT group, P = 0·003). Adherence to ACEI was also higher in the PCI group, but the difference was not statistically significant (77·6% vs. 69·3%, P = 0·17). By logistic regression analysis, belonging to the PCI group was an independent predictor of medication adherence [B = 2·20 (1·06-4·50), P = 0·03)]. What is new and conclusion In the present study we demonstrated that adherence to evidence-based medication therapies in patients treated with PCI is significantly higher than in patients treated with OMT alone. Medication adherence should be followed carefully in CAD patients treated with OMT. © 2013 John Wiley & Sons Ltd.