Use of evidence-based medicine (EBM) improves outcomes after acute coronary syndromes (ACS), yet patients often discontinue prescribed therapies after discharge. Although such discontinuation is well documented, patients' reasons for medication discontinuation have not been reported. MAINTAIN is a longitudinal follow-up registry of CRUSADE/ACTION, which enrolled patients during an ACS hospitalization from January 2006 to September 2007. All discharge medications were obtained from hospital charts. Patients were interviewed by telephone 3 months after discharge to determine if EBM classes prescribed at discharge were continued (aspirin, clopidogrel, beta blockers, angiotensin-converting enzyme inhibitors/angiotensin receptor blockers, and lipid-lowering medications). If discontinuation occurred, patients were asked if it was with provider knowledge/input or not (self-discontinuation). A multivariable logistic regression model was performed to identify factors associated with self-discontinuation of prescribed EBM. Of the 1,077 patients interviewed, 1,006 (93.4%) were discharged on aspirin, 816 (75.8%) on clopidogrel, 982 (91.2%) on beta blockers, 745 (69.2%) on angiotensin-converting enzyme inhibitors/angiotensin receptor blockers, and 968 (89.9%) on lipid-lowering medications. At 3-month follow-up, 304 patients (28.2%) had discontinued > or =1 of these prescribed EBM classes. Although many reported provider involvement, most discontinuation (61.5%) was self-determined. Factors independently associated with self-discontinuation were no pharmacy coinsurance, increasing number of medications, not using reminder tools (e.g., pillbox), lower education, and dialysis. In conclusion, 1/3 of patients with ACS discontinue > or =1 of their prescribed EBMs within 3 months of hospital discharge, and most of this discontinuation is without provider involvement. Patient education, better prescription drug coverage, and reminder strategies may improve use of EBMs at 3 months after discharge from ACS admission.