A lack of aspirin effect on platelets after a myocardial infarction (MI) is associated with poor health outcome. This lack of effect may be due to biological resistance to aspirin or due to nonadherence (the patient is not taking the aspirin, hence it has no effect). Determining which of these factors predicts poor outcome would inform potential intervention strategies.Aspirin effect on platelets was assessed in a cohort of MI survivors who were divided into three groups: group A ("adherent"), patients whose platelets were affected by aspirin; group B ("nonadherent"), patients whose platelets showed no aspirin effect and who admitted in an interview that they were not taking their medications; and group C (potentially biologically resistant to aspirin), patients whose platelets showed no aspirin effect but maintained that they were taking their aspirin. Two health outcome measures (death, reinfarction, or rehospitalization for unstable angina; or admission for any cardiovascular causes) were assessed 12 months after enrollment.Seventy-three patients were enrolled and classified into groups A ("adherent," 52 patients), B ("nonadherent," 12 patients), and C ("potentially aspirin resistant," 9 patients). Adverse events and readmission were more common in the nonadherent group (B)-42% and 67%, respectively, when compared with the adherent group (A)-6% and 11%, and with the potentially biologically resistant group (C)-11% and 11%.Nonadherence is a significant mediator of poor outcome. It is important to evaluate whether or not patients are taking their medications in clinical settings and in studies that evaluate the effect of prescribed medications.