Secondary prevention programmes for Coronary Heart Disease (CHD) aim to reduce cardiovascular risks and promote health in people with heart disease. Though programmes have been associated with health improvements in study populations, access to programmes remains low, and quality and effectiveness is highly variable. Current guidelines propose significant modifications to programmes, but existing research provides little insight into why programme effectiveness varies so much. Drawing on a critical realist approach, this article argues that current research has been based on an impoverished ontology, which has elements of positivism, does not explore the social determinants of health or the effects on outcomes of salient contextual factors, and thereby fails to account for programme variations. Alternative constructivist approaches are also weak and lacking in clinical credibility. An alternative critical realist approach is proposed that draws on the merits of subjectivist and objectivist approaches but also reflects the complex interplay between individual, programme-related, socio-cultural and organizational factors that influence health outcomes in open systems. This approach embraces measurement of objective effectiveness but also examines the mechanisms, organizational and contextual-related factors causing these outcomes. Finally, a practical example of how a critical realist approach can guide research into secondary prevention programmes is provided.