Participants of a randomised trial may differ from eligible non-participants as a result of selection. We studied the distribution of prognostic factors and survival in eligible patients of a multi-centre trial of long-term oral anticoagulant treatment after myocardial infarction. All hospital survivors of myocardial infarction in one participating clinical centre of a multi-centre, randomised, double-blind, placebo-controlled trial of long-term anticoagulant treatment after myocardial infarction were screened for entry criteria. Subsequently, prognostic factors and survival of participants were compared with eligible but not randomised patients. The 350 participants were younger and were more often of male gender and more often smokers compared with 587 non-participants. Non-participants had more frequently suffered a previous myocardial infarction and were treated more often with diuretics and ACE-inhibitors, suggesting a higher proportion of patients with chronic heart failure in this group. Age, previous myocardial infarction and the use of diuretics at discharge were independent predictors of mortality, consent showed no association. Our findings indicate that participants of a clinical trial have a better prognosis during the first years following myocardial infarction compared to eligible non-participants as a result of a higher prevalence of cardiovascular risk factors associated with mortality in the non-participants.